Thoracic back pain and suspected thoracic myelopathy

ACR AC Myelopathy (2021); ACR AC Low Back Pain (2021); NICE NG234

Two things make this different from low back pain. Thoracic pain carries a materially higher pre-test probability of malignancy, infection and fracture — degenerative thoracic disease is comparatively uncommon, so the symptom is closer to a red flag in its own right — and if there are myelopathic features the lesion is frequently not at the level the patient points to, which is an argument for imaging the whole cord rather than the painful segment.

Pain localised to the thoracic spine or radiating in a band around the chest or abdomen, with or without gait disturbance, leg stiffness, a sensory level or sphincter change. The card covers the subacute and elective presentation; an acute evolving cord syndrome belongs on the inflammatory cord lesion or metastatic cord compression pathways.

Referenced decision support — confirm against your local protocol.

Decision support, not a directive. Protocols and timings shown are typical published ones — your local protocol takes precedence, and the vetting radiologist decides.

The request in front of you

Everything is optional. Leave a field alone and the answer assumes nothing — the verdict updates as you go.

Study requested

What the referrer actually asked for. It is evidence of intent, not a constraint on the right answer.

Contrast as written

What the request form says, not what it should say.

Age

Pick a band, or type an exact age if it matters.

years

Pregnancy status

Time since onset

Decides thrombolysis and thrombectomy windows, testicular salvage, and whether an ischaemic limb is still salvageable.

Fever or sepsis

Immunosuppressed

Changes the differential rather than the modality.

Red flags

These accumulate rather than exclude. Back pain with one of these is a different question from back pain.

Known cancer

Back pain in a cancer patient is a different question, and staging phase changes which contrast phases are needed.

Cord features

Weakness, a sensory level, hyperreflexia, sphincter or gait disturbance. This changes how much of the spine is imaged, not which scanner is used — a sensory level says the cord is involved but not reliably where, and the lesion is often several segments higher.

Renal risk factors

The question a vetter can actually answer from the request. An explicit “none known” is a real answer, and it removes checks rather than deferring them.

Previous contrast reaction

The pathway — tap anything already done

Marking a study complete moves the answer on. A patient arrives partway through a pathway far more often than at the start of one.

Accept as requested

MRI Thoracic Spine — routine

MRI Thoracic Spine

What we'd amend, and why

MRI is the only study that shows all four of the things that actually cause thoracic pain worth imaging — marrow, disc, epidural space and cord — and the marrow is the one that matters most, because the thoracic spine is where metastatic and myelomatous deposits concentrate and where a deposit is invisible on radiographs until roughly a third of the trabecular bone has gone. The sagittal fluid-sensitive fat-suppressed sequence is the workhorse: it separates an acute oedematous vertebral fracture from an old healed one, shows marrow replacement before there is any bony collapse, and detects the disc and endplate signal change of early discitis. Regional coverage is defensible when the pain is localised and there are no cord features. The technical caveats are real and belong in the request: cardiac and respiratory motion degrade the mid-thoracic cord more than any other spinal segment, and level counting on a thoracic-only study is the single most consequential error made here.

Nothing needs resolving before this goes ahead. Routine checks below are owned downstream.

How was this decided?

pathwayadult — Adults — thoracic pain with or without cord features
rulerule-mr-device-screening — MR safety screening for implants and foreign bodies; checked by Radiographer at the scanner

Decision support only. Local protocol takes precedence.

Handled at the scanner(1)nothing for you to do

Settled and owned downstream. Each returns to a radiologist only on the stated trigger.

MR safety screening for implants and foreign bodies
Complete the MR safety questionnaire, verify implant labelling and its stated conditions of use against this scanner and this protocol, and ensure no ferromagnetic object enters Zone IV.
Radiographer at the scannerBefore the scan
Flags back if: An implant or retained foreign body that is MR Unsafe, unlabelled, or cannot be identified; or an MR Conditional device whose stated conditions this scanner or the requested protocol cannot satisfy; or a credible unexcluded intraocular metallic foreign body history.

Worth asking the referrer (3)

None of these hold the request up. They sharpen the protocol or the plan that follows.

Are there myelopathic features — gait disturbance, leg stiffness or spasticity, brisk reflexes, a sensory level, or bladder or bowel change?
It decides the coverage rather than whether to scan. A myelopathic patient needs the whole cord, because the sensory level is an unreliable localiser and a cervical lesion regularly presents with leg-predominant symptoms; a patient with localised thoracic pain and no cord signs is properly served by a regional study.
Is there a history of malignancy, unexplained weight loss, night pain, fever or a source of bacteraemia, intravenous drug use, immunosuppression, osteoporosis or steroid use, or recent trauma?
These raise the pre-test probability that is already higher in the thoracic spine than the lumbar, and several of them redirect the card entirely — to metastatic cord compression, to spinal infection, or to the thoracolumbar trauma pathway.
How quickly are the neurological symptoms progressing — hours, days or months?
An acute or rapidly evolving cord syndrome is a same-day pathway with a compressive question at the front of it, and is answered by a different card. This one covers the subacute and elective presentation.

Pathways

Big forks are separate pathways; the first whose conditions match is the one used.

Fever, immunosuppression or suspected infection — image for discitis

Role	Study & protocol	Why this answers the question
First line	MRI Whole Spine MRI Spine — suspected infection / discitis usually appropriate	Thoracic pain with fever or in an immunosuppressed patient is discitis, vertebral osteomyelitis or an epidural abscess until the imaging says otherwise, and the regional unenhanced study this card otherwise returns is the wrong study twice over. Gadolinium is what separates an enhancing phlegmon from a drainable collection and defines the epidural component that decides surgery, and whole-spine sagittal coverage is what finds the non-contiguous second level that is characteristic of tuberculous disease. Immunosuppression also widens the organism list and blunts the inflammatory response, so the clinical threshold that would be reassuring in anyone else is not reassuring here.

Cord features evolving over hours — a same-day compressive question

Role	Study & protocol	Why this answers the question
First line	MRI Whole Spine MRI Whole Spine — suspected metastatic cord compression usually appropriate	This card covers the subacute and elective presentation, and cord signs that have appeared within the last day or two are neither. The tempo moves the compressive question to the front — metastatic deposit, epidural collection, disc or haematoma — and the deficit present when treatment starts is broadly the deficit the patient keeps, so this is a study today rather than a slot next week. Whole-cord coverage for the reason the card gives everywhere else: sensory levels localise poorly, the responsible lesion is frequently several segments higher, and a cervical lesion routinely presents with leg-predominant symptoms and no neck pain.
Problem solving	CT Thoracic Spine CT Thoracic Spine — Unenhanced, Thin Section	Where MRI is contraindicated or genuinely undeliverable today. It shows bone destruction, retropulsion and canal calibre and it does not show the cord, so a normal CT in an acutely myelopathic patient excludes nothing that matters and the MRI question stays open — transfer rather than substitution is usually the right response.

Weight loss, night pain or a stated malignancy red flag — whole spine

Role	Study & protocol	Why this answers the question
First line	MRI Whole Spine MRI Whole Spine — suspected metastatic cord compression usually appropriate	The thoracic spine is where metastatic and myelomatous deposits concentrate, and the prior probability that made this card different from low back pain in the first place is exactly what these red flags raise further. Regional coverage cannot exclude the diagnosis it is most likely to find: deposits are frequently multilevel and non-contiguous, and the level that determines the radiotherapy field or the surgical construct is regularly not the level that hurts. Weight loss or night pain without a known primary is a weaker signal than a cancer history and still enough to buy the extra sagittal coverage in the same attendance.

Myelopathic features — image the whole cord

Role	Study & protocol	Why this answers the question
First line	MRI Whole Spine MRI Whole Spine — suspected metastatic cord compression	A sensory level tells you that the cord is involved; it does not reliably tell you where. The responsible lesion sits above the clinical level often enough that imaging only the painful segment is the recognised way to miss it, and a cervical lesion in a patient complaining solely of interscapular pain is the case that ends badly. Whole-spine sagittal screening costs a few minutes on the same visit and removes the question entirely, with focused axials through whatever it finds.
Problem solving	CT Thoracic Spine CT Thoracic Spine — Unenhanced, Thin Section	Where MRI is contraindicated or cannot be obtained in a clinically useful timeframe, CT shows bone destruction, retropulsion and canal calibre. It does not show the cord, so a normal CT in a myelopathic patient excludes nothing that matters and the MRI question stays open.

Known malignancy — this is a spinal metastasis question

Role	Study & protocol	Why this answers the question
First line	MRI Whole Spine MRI Whole Spine — suspected metastatic cord compression usually appropriate	New thoracic pain in a patient with known cancer is a red flag by definition, and the thoracic spine is the commonest site of spinal metastatic disease — so the study has to be able to exclude the diagnosis it is most likely to find. Regional coverage cannot: metastatic deposits are frequently multilevel and non-contiguous, referred pain localises poorly, and the level that determines the radiotherapy field or the surgical construct is regularly not the level that hurts. NICE asks for whole-spine MRI within a week of suspected spinal metastases and within 24 hours where there are neurological signs. Imaging only the painful segment here treats the symptom and leaves the lesion that will paralyse the patient outside the field of view.

This arm exists to refuse the regional answer. Move to the metastatic spinal cord compression pathway for the urgency banding and the treatment sequence.

Adults — thoracic pain with or without cord features

Matches your inputsDefault

Role	Study & protocol	Why this answers the question
First line	MRI Thoracic Spine MRI Thoracic Spine — routine usually appropriate	MRI is the only study that shows all four of the things that actually cause thoracic pain worth imaging — marrow, disc, epidural space and cord — and the marrow is the one that matters most, because the thoracic spine is where metastatic and myelomatous deposits concentrate and where a deposit is invisible on radiographs until roughly a third of the trabecular bone has gone. The sagittal fluid-sensitive fat-suppressed sequence is the workhorse: it separates an acute oedematous vertebral fracture from an old healed one, shows marrow replacement before there is any bony collapse, and detects the disc and endplate signal change of early discitis. Regional coverage is defensible when the pain is localised and there are no cord features. The technical caveats are real and belong in the request: cardiac and respiratory motion degrade the mid-thoracic cord more than any other spinal segment, and level counting on a thoracic-only study is the single most consequential error made here.
First line	MRI Whole Spine MRI Whole Spine — suspected metastatic cord compression	The right extent whenever the level is not established: myelopathic signs, a sensory level, an unexplained progressive gait disturbance, or a suspicion of multifocal disease. Two facts drive it. A sensory level is a poor localiser and the responsible lesion is often several segments above it, and a cervical cord lesion frequently presents with leg-predominant symptoms and no neck pain, so a thoracic-only study in that patient can be entirely normal while the disease sits above the field of view. Acquiring the extra coverage in the same attendance costs sagittal sequences and a few minutes; discovering the need afterwards costs a second appointment, and in a progressive myelopathy the interval is not neutral.
Problem solving	CT Thoracic Spine CT Thoracic Spine — Unenhanced, Thin Section	A complement to MRI rather than a substitute for it, and the distinction is not academic: CT cannot show cord signal change, cannot reliably show an epidural soft-tissue mass, and cannot detect marrow infiltration before the bone has been destroyed, so a normal CT in a patient with thoracic red flags is not a negative result. Where it earns its place is bony detail once the MRI has been done or cannot be — characterising a vertebral collapse and its retropulsion, showing whether a thoracic disc is calcified, which changes the surgical approach, and planning instrumentation. It is also the fallback when MRI is genuinely contraindicated, and in that situation the report should say plainly what the study could not exclude. Before scanning, check for an existing chest or trauma acquisition: thin reformats from a dataset already on the system usually answer the bony question with no further exposure.

Radiographs are deliberately off this ladder. They cannot exclude the malignancy, infection or cord pathology that make thoracic pain worth imaging, and a normal thoracic radiograph in a patient with red flags is a false reassurance bought with marrow and breast dose.
If the pain is acute and follows trauma, or the patient has osteoporosis and a low-energy mechanism, the thoracolumbar trauma pathway is the right card.
Contrast is not part of the routine study. It is added when infection, tumour or inflammatory cord disease is the question, which is usually a different card.

Pitfalls

Applying the low back pain reflex to thoracic pain. Degenerative thoracic disease is comparatively uncommon, the pre-test probability of malignancy, infection and fracture is higher, and "wait six weeks" is the wrong default in this region.
Imaging only the painful segment in a myelopathic patient. Sensory levels are unreliable localisers and the responsible lesion is frequently several segments higher — sometimes in the cervical cord, in a patient with no neck pain at all.
Miscounting vertebral levels on a thoracic-only study. It is the most consequential technical error here, it can result in surgery at the wrong level, and it is preventable with a localiser that reaches a countable landmark or a prior radiograph.
Accepting a normal CT as excluding the diagnosis. CT cannot show cord signal change or marrow infiltration, and a reassuring CT in a patient with thoracic red flags is actively misleading.
Mistaking cerebrospinal fluid pulsation artefact for an intradural lesion, which is commoner in the thoracic canal than anywhere else in the spine.
Missing a spinal dural arteriovenous fistula in an older patient with progressive myelopathy: cord oedema over several segments with dilated perimedullary flow voids, and treatable if it is recognised.
Reporting the disc and forgetting the marrow. The commonest serious finding in this region is a vertebral deposit, and it is seen on the fluid-sensitive and T1 sagittals rather than on the axial disc images.
Attributing band-like chest or abdominal pain to a visceral cause without considering a thoracic radicular origin, and the reverse — the thoracic dermatomes send a lot of patients through cardiac and abdominal workups first.

Priors — what to pull first

Any prior cross-sectional imaging that included the thoracic spine — a staging CT, a CT pulmonary angiogram, an abdominal CT — frequently already shows or excludes the vertebral lesion in question, and the sagittal reformats are often better than anyone realises.
A previous study is what separates an old healed osteoporotic collapse from an acute or pathological one, which is otherwise among the harder calls in this region.
Where the patient has had thoracic surgery or instrumentation, the operative note frames both the expected artefact and which level is being asked about.

What makes a good request

The epidemiology does the work here. Symptomatic thoracic degenerative disease is uncommon because the rib cage splints the segment, so a patient complaining of thoracic pain has a higher prior probability of a vertebral metastasis, a myeloma deposit, discitis or an osteoporotic or pathological collapse than the far commoner lumbar presentation. Applying the "no imaging for six weeks" reflex from low back pain to this region is the characteristic error.
Myelopathic features change the coverage, not just the urgency. Sensory levels localise poorly and the responsible lesion is frequently several segments higher than the clinical level, and a cervical cord lesion routinely presents with leg-predominant symptoms and no neck pain at all — which is why a thoracic-only study in a myelopathic patient can be normal while the disease sits above the field of view.
The ACR Appropriateness Criteria on myelopathy make MRI the first-line study for cord symptoms, and NICE NG234 asks for whole-spine imaging where the level of disease is not established, which is exactly the myelopathic patient.
Level counting is a real hazard in this region and has surgical consequences. A thoracic-only acquisition gives the reporter no reliable landmark, so a localiser that includes the cervical or lumbar spine, or a prior radiograph, should be available before a level is committed to a report.
Thoracic disc herniations are frequently calcified and often broad-based and central, which changes the operative approach and is one of the few things CT genuinely adds once the MRI is done.
A spinal dural arteriovenous fistula is the diagnosis to keep in mind in an older patient with progressive myelopathy and no compressive lesion: cord oedema over several segments with dilated perimedullary flow voids, treatable, and routinely missed for months.

Scoring this once it is done

The classification and risk tools this question ends in.

ASIA Impairment Scale (AIS) with Frankel Mapping Meyerding Grading (Spondylolisthesis)SINS (Spinal Instability Neoplastic Score)Bone-RADS (Bone Lesion Scoring)RECIST 1.1 Tumor Response

How these studies are acquired

Contrast, phases and timing for every study on the pathways above.

MRI Spine — suspected infection / discitis MRI Whole Spine — suspected metastatic cord compression CT Thoracic Spine — Unenhanced, Thin Section MRI Thoracic Spine — routine

Confirm locally

MRI Thoracic Spine — routine: timings are typical — confirm against local protocol.
Timings, contrast volumes and rates above are typical published values. Your department's protocol, scanner and patient population decide the actual numbers.
Safety thresholds and premedication policy follow local policy where it differs from the cited guidance.

References

Implemented from the cited published sources. Educational and workflow support only; confirm against current guidelines and local policy before clinical use.