MRI Orbits for Thyroid Eye Disease / Graves' Orbitopathy

Version 1.0 — August 2026

1. Executive Summary

1.1 Added Value over the Generic Protocol

Thyroid eye disease (TED, also called Graves’ orbitopathy) is the most common orbital disease encountered in routine practice, and its MRI assessment depends on genuinely quantitative measurement — extraocular muscle diameter, muscle T2/STIR signal intensity relative to a reference tissue, and apical crowding at the orbital apex — rather than the qualitative mass-lesion characterisation that occupies most of the rest of this cluster. This child page documents that quantitative measurement convention, the imaging-based activity assessment that complements (but does not replace) the clinical activity score used in endocrinology practice, and — the single most clinically urgent task this protocol performs — the specific imaging features that identify dysthyroid optic neuropathy (DON), a sight-threatening complication that can be present with a deceptively unremarkable clinical examination.

1.2 Limits of the Dedicated Protocol

MRI-based activity markers (T2/STIR signal intensity ratio, T2 relaxation time) correlate meaningfully with the clinical activity score but are not a validated, stand-alone replacement for it; current management guidelines continue to base treatment decisions on the combined clinical and imaging picture rather than imaging criteria alone. Similarly, while specific MRI features (apical crowding in particular) carry a substantially increased likelihood of dysthyroid optic neuropathy, the diagnosis remains a clinical one, supported rather than made by imaging, and a technically unremarkable study in a patient with a compatible clinical picture should not be used to exclude DON outright.

2. Clinical Context

2.1 Clinical Presentation

TED presents with a spectrum from mild (lid retraction, mild soft-tissue signs) through moderate-to-severe (significant proptosis, diplopia from restrictive myopathy, exposure keratopathy) to sight-threatening disease (dysthyroid optic neuropathy or corneal breakdown), most often — though not always — in the context of established or newly diagnosed Graves’ hyperthyroidism, with a minority of cases occurring in euthyroid or hypothyroid autoimmune thyroid disease. Bilateral, though frequently asymmetric, involvement is characteristic and itself a useful distinguishing feature from many of the unilateral mass lesions discussed in the companion Orbital Tumour and Mass Lesion child page.

2.2 The Two-Phase Natural History — Why Activity Assessment Matters

TED characteristically follows an active inflammatory phase (months to a few years, during which immunosuppressive treatment is genuinely effective) followed by a burnt-out, fibrotic inactive phase (during which the same immunosuppressive treatments are largely ineffective, and management shifts toward rehabilitative surgery); correctly identifying which phase a given patient is in — Section 2.1 of the parent EUGOGO clinical practice guideline explicitly frames this as fundamental to treatment choice — is the central clinical question this protocol’s activity-assessment role (Section 5.2) is designed to support.

2.3 Differential Diagnosis (Clinical)

Bilateral extraocular muscle enlargement with a compatible clinical thyroid history is rarely diagnostically ambiguous, but idiopathic orbital inflammation (myositis subtype) and, less commonly, IgG4-related orbital disease or orbital lymphoma can each produce extraocular muscle enlargement mimicking TED — addressed further in Section 5.5 — and should specifically be considered when the clinical thyroid context is genuinely absent, when involvement is markedly asymmetric or unilateral, or when pain is a prominent early feature (atypical for classic TED, which is more often associated with a foreign-body/pressure sensation than frank pain).

3. Indications, Timing, and Patient Selection

3.1 When the Dedicated Protocol Is Indicated

New or clinically evolving proptosis, lid retraction, or diplopia in a patient with known or suspected thyroid dysfunction; any TED patient in whom the activity phase (Section 2.2) is clinically uncertain and would change management; and — with the urgency described in Section 3.3 — any TED patient with symptoms or signs raising concern for dysthyroid optic neuropathy are all indications for this dedicated protocol.

3.2 Baseline and Treatment-Response Imaging

Because activity-phase treatment decisions (Section 2.2) and their monitoring both benefit from objective, comparable imaging, a well-characterised baseline study — using consistent sequence parameters, and ideally the same measurement convention (Section 5.1) at each timepoint — supports both the initial activity assessment and, where immunosuppressive treatment is given, subsequent objective assessment of treatment response, complementing the clinical activity score used in parallel.

3.3 Red Flags Modifying Urgency — Dysthyroid Optic Neuropathy

Progressive visual loss, new or worsening colour desaturation, or a new relative afferent pupillary defect in a TED patient are the clinical hallmarks of dysthyroid optic neuropathy and constitute a genuine ophthalmological emergency; imaging in this setting should be expedited, and the report should explicitly and promptly address the specific DON-associated features documented in Section 5.3, since DON is a leading cause of preventable, treatment-responsive visual loss in TED when identified promptly.

4. Dedicated Protocol Design

4.1 Mandatory Core Sequences

The table below lists the complete mandatory protocol for thyroid eye disease assessment — the seven generic-protocol core sequences (1-7) plus the one dedicated addition (8) detailed in Section 4.3. See Section 4.2 for the full comparison against the generic protocol.

4.2 Protocol Delta vs the Generic Protocol

4.3 Sequence-by-Sequence Utility, Including the Fat-Suppression Technique Case for This Indication

Sequences 1-2 and 8 (fat-suppressed T2) — muscle signal and diameter, the two central quantitative measurements of this protocol. Signal intensity within the extraocular muscles on fat-suppressed T2/STIR imaging reflects tissue water content from active inflammatory oedema, and a landmark original study directly correlating extraocular muscle signal intensity on STIR-sequence MRI with the Mourits clinical activity score found that the area of highest signal intensity within the most inflamed muscle, and the average cross-sectional signal intensity at that point, both reliably correlated with clinical activity and tracked its change over time — establishing the basic principle underlying Section 5.2’s activity assessment. This is also the specific reason the parent master page’s Section 4.3 fat-suppression reasoning matters more, not less, in this particular child protocol: an orbit-specific comparative study found Dixon-T2WI at 3T outperformed conventional fat-suppression technique specifically in thyroid eye disease, with more uniform, higher-SNR fat suppression translating directly into more reliable muscle and lacrimal gland signal assessment — precisely the quantitative measurement this sequence pair exists to support, meaning a poorly or unevenly fat-suppressed T2 sequence does not merely look worse here, it can genuinely distort the activity assessment itself.

Sequence 3 (non-fat-suppressed T1) — baseline anatomy and myopathy-mimicking mass exclusion. Beyond its generic role, this sequence supports confirming that muscle enlargement is genuinely myopathic (diffuse, fusiform, tendon-sparing, per Section 5.1) rather than reflecting a discrete intramuscular mass or other structural lesion, a distinction occasionally genuinely necessary in atypical presentations.

Sequences 4-5 (post-contrast T1) — conditional, not routine, for this indication. Unlike the tumour and mass lesion protocol, standard TED activity and severity assessment does not routinely depend on enhancement pattern; contrast is reserved specifically for atypical presentations where the differential extends beyond classic TED (Section 2.3), or where a specific surgical-planning question requires it, per Section 4.5.

Sequence 6 (DWI/ADC, apex-focused) — dysthyroid optic neuropathy risk stratification. A recent, large comparative study of MRI features in dysthyroid optic neuropathy found elevated apparent diffusion coefficient at the orbital apex was independently associated with substantially increased odds of DON, alongside apical crowding — establishing genuine, quantifiable diagnostic value for diffusion imaging specifically at the apex in this indication, distinct from DWI’s more general cellularity-marker role in the companion Orbital Tumour and Mass Lesion child page.

Sequence 7 (whole-brain T2/FLAIR through the chiasm) — excluding a genuinely separate cause of visual dysfunction. Since DON is a diagnosis that specifically requires excluding an unrelated intracranial cause for any visual field or acuity change, this sequence’s chiasm-and-beyond coverage retains its full generic-protocol importance in this indication, not merely as a formality.

4.4 Measurement Plane Standardisation for Sequence 8

Because extraocular muscle diameter measurement (Section 5.1) depends on genuinely reproducible slice positioning, Sequence 8 should be planned specifically perpendicular to the long axis of the optic nerve at a consistent, described anatomical level (typically the mid-orbit, at or near the point of maximal muscle diameter) — an inconsistent or poorly reproduced measurement plane between serial studies introduces genuine measurement variability that can be mistaken for true interval change (Section 3.2).

4.5 Contrast Strategy

Gadolinium contrast is not routinely required for standard TED activity/severity assessment, since the central measurements this protocol depends on (muscle diameter, T2/STIR signal intensity, apical crowding) are all non-contrast findings; it should be added specifically where the clinical differential genuinely extends beyond classic TED (Section 2.3), or for dedicated surgical planning in complex orbital decompression cases, using standard macrocyclic GBCA dosing with Dixon or spectral/SPIR fat suppression per the parent master page’s Section 4.3 reasoning — never STIR.

4.6 Sequence Matching to Clinical Question

5. MRI Semiotics of Thyroid Eye Disease

5.1 Extraocular Muscle Involvement — Pattern and Measurement

TED characteristically produces diffuse, fusiform enlargement of the extraocular muscle bellies with relative sparing of the tendinous insertions — a specific morphological pattern that itself helps distinguish TED from idiopathic orbital myositis, which more typically involves the tendon as well as the belly (Section 5.5). Involvement follows a recognised, though not invariable, order of frequency (classically remembered as inferior rectus most commonly affected, followed by medial, superior, then lateral rectus), and muscle diameter is measured on the standardised coronal plane established in Section 4.4, compared against locally or literature-established normal reference ranges, with bilateral, side-by-side comparison — given the characteristic bilateral-but-asymmetric pattern — adding further diagnostic confidence beyond an isolated absolute measurement.

5.2 Imaging-Based Activity Assessment

Signal intensity ratio — the fat-suppressed T2/STIR signal intensity of the most involved extraocular muscle expressed relative to a reference tissue (commonly ipsilateral temporalis muscle or white matter) — is the principal imaging-based activity marker, with higher ratios associated with active, inflammatory-phase disease and correlating with the clinical activity score, as established in the foundational STIR/CAS correlation literature. This imaging-based activity assessment is intended to complement, not replace, the clinical activity score in guiding the treatment-phase decision described in Section 2.2, and should be reported as a contributing piece of evidence rather than an independent diagnostic verdict.

5.3 Dysthyroid Optic Neuropathy — the Critical Complication

Apical crowding — enlarged extraocular muscle bellies compressing the optic nerve at the comparatively confined orbital apex — is the dominant mechanism and single most important imaging feature associated with dysthyroid optic neuropathy; a recent large comparative study found apical crowding carried a markedly increased odds ratio for DON, together with elevated apex-level ADC, extraocular muscle enlargement, and peri-muscular fat/muscle signal intensity, all significantly more frequent in DON than non-DON cases. Because DON can occur with a comparatively modest degree of proptosis (the crowded, non-expansile orbital apex providing less room to accommodate muscle enlargement than the more capacious mid-orbit), the absence of severe proptosis should never be used to argue against active apical crowding assessment when DON is a clinical concern.

5.4 Severity and Extent — Additional Findings

Beyond muscle and apex assessment, orbital fat expansion, lacrimal gland involvement (enlargement, sometimes with a similar signal-intensity-based activity assessment to that described for muscle in Section 5.2), and globe subluxation/proptosis severity all contribute to the overall severity picture, documented per the structured reporting template in Section 6.1, complementing rather than substituting for the muscle- and apex-focused core assessment.

5.5 Differential Diagnosis on Imaging

Idiopathic orbital myositis characteristically involves the tendinous insertion as well as the muscle belly (in contrast to TED’s characteristic tendon-sparing pattern, Section 5.1), is more often unilateral, and is more frequently associated with pain; IgG4-related orbital disease and orbital lymphoma (addressed in the companion Orbital Tumour and Mass Lesion child page) can each occasionally produce extraocular muscle enlargement mimicking TED, but generally lack the specific combination of bilateral, tendon-sparing, fusiform muscle enlargement with a compatible thyroid clinical context that defines classic TED.

5.6 Mimickers and Pitfalls

Because some baseline extraocular muscle diameter and signal intensity variation exists in the normal population, an isolated measurement modestly above a published reference range, without the characteristic bilateral (though asymmetric), tendon-sparing morphology and compatible clinical thyroid context, should be interpreted cautiously rather than treated as diagnostic of TED in isolation — exactly the same general interpretive caution already established for other quantitative imaging signs throughout this cluster.

6. Reporting Framework

6.1 Structured Reporting Template

Extraocular muscle involvement: which muscles, diameter measurements (Section 5.1), bilateral comparison. Muscle morphology: fusiform, tendon-sparing pattern explicitly confirmed or noted as atypical. Signal intensity/activity assessment: qualitative or, where measured, quantitative signal intensity ratio (Section 5.2). Apical crowding: explicitly assessed and graded, with specific comment whenever DON is a clinical concern. DWI/ADC at the apex: explicitly reported when dysthyroid optic neuropathy is a clinical concern (Section 4.3). Other findings: orbital fat expansion, lacrimal gland involvement, proptosis severity. Comparison with prior imaging: stable/progressed/improved, explicitly stated when a baseline exists.

6.2 Mandatory Reporting Elements

Every report should explicitly address apical crowding status, even when DON is not the primary clinical question, given how much apical crowding’s presence changes management urgency, and should explicitly distinguish an activity assessment (Section 5.2) from a severity assessment (Section 5.4) rather than presenting a single, undifferentiated impression that conflates the two genuinely distinct clinical questions.

6.3 Critical/Actionable Findings

Significant apical crowding, particularly combined with elevated apex-level ADC (Section 5.3), in a patient with clinical concern for dysthyroid optic neuropathy is the single most urgent, directly actionable finding category in this protocol and should be communicated promptly, given DON’s status as a genuinely treatable but time-sensitive cause of visual loss.

6.4 Common Reporting Errors

Reporting muscle enlargement without explicit diameter measurement or comparison against a reference range; conflating activity assessment and severity assessment into a single undifferentiated description; omitting explicit comment on apical crowding in a study performed for a DON-related clinical question; and treating an isolated, mildly abnormal measurement as diagnostic of TED without considering the differential in Section 5.5.

7. Technical Pitfalls

7.1 Measurement Plane Reproducibility

As emphasised in Section 4.4, inconsistent slice positioning between serial studies is a genuine, avoidable source of apparent interval change that is actually measurement artefact — a particularly consequential pitfall given how directly treatment decisions in this indication depend on genuine, correctly-identified interval change.

7.2 Sequence-Specific Technical Considerations

Because signal intensity ratio (Section 5.2) depends on comparison against a reference tissue, inconsistent reference-region placement (e.g., inconsistent temporalis muscle sampling location) between the muscle-of-interest measurement and the reference measurement introduces a further, avoidable source of quantitative variability distinct from the imaging technique itself.

7.3 When the Generic Protocol Alone Is Insufficient

A study performed using the generic protocol’s default approach, without the dedicated, reproducible measurement-plane coronal sequence (Section 4.4) and without apex-focused DWI when DON is a concern, risks under-characterising exactly the two pieces of information (quantitative muscle/activity assessment, apical crowding/DON risk) that are this protocol’s central purpose.

8. MRI Technologist Pearls

8.1 Sequence Planning for Reproducible Measurement

Plan Sequence 8’s coronal plane deliberately and consistently, ideally referencing the same anatomical landmark description used on any prior study for the same patient, rather than a freehand or purely visually-judged slice position — this single planning decision has an outsized effect on this protocol’s overall clinical value.

8.2 Coil and Positioning Considerations

As in the generic protocol, confirm genuinely symmetric bilateral coverage, which matters especially here given how central bilateral comparison is to both the diameter measurement (Section 5.1) and activity assessment (Section 5.2) central to this protocol.

8.3 Fast Salvage Protocol

If time is genuinely constrained, prioritise the coronal fat-suppressed T2 measurement sequence (Sequence 8) and, where DON is a concern, the apex-focused DWI (Sequence 6) over post-contrast sequences, which are conditional rather than routine for this indication (Section 4.5).

8.4 Disease-Specific Common Avoidable Errors

Acquiring the coronal T2 measurement sequence with inconsistent plane positioning relative to prior studies; omitting apex-focused DWI when DON is a genuine clinical concern; obtaining contrast routinely despite it not being required for standard activity/severity assessment in this indication; and failing to explicitly assess and report apical crowding status even when not specifically requested.

9. Quality Control Checklist

• Coronal fat-suppressed T2 measurement sequence (Sequence 8) confirmed acquired with a consistent, reproducible measurement plane, referenced against any prior study.
• Bilateral, symmetric coverage confirmed for direct side-to-side comparison.
• Apex-focused DWI/ADC confirmed acquired whenever dysthyroid optic neuropathy is a clinical concern.
• Apical crowding explicitly assessed and reported, independent of whether DON was the primary indication.
• Activity assessment and severity assessment confirmed reported as distinct, separately-addressed clinical questions.
• Contrast confirmed genuinely indicated (Section 4.5), not obtained by routine default.

11. Evidence Gaps and Ongoing Debate

• Imaging-based activity assessment is not yet a formally validated stand-alone replacement for the clinical activity score. As emphasised throughout this page, current guidelines continue to base treatment decisions on the combined clinical and imaging picture; the precise, universally-adopted quantitative imaging threshold for “active” versus “inactive” disease remains less standardised than the underlying correlation itself.
• Quantitative T2 relaxometry, while technically promising for more reproducible activity assessment, remains less widely validated and standardised across institutions than the qualitative/semi-quantitative signal-intensity-ratio approach that represents current mainstream practice (Section 10).
• The precise ADC and apical-crowding thresholds most predictive of dysthyroid optic neuropathy are drawn from a single, albeit large and recent, comparative cohort, and further independent validation across different institutions and imaging protocols would strengthen confidence in the specific quantitative cut-offs used in Section 5.3.

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

C. Important Prospective / Original Studies

D. Technical MRI Papers

Represented by the original studies already listed under Category C; a separate, non-duplicative Category D entry is not populated to avoid citing the same sources twice.

E. Landmark Historical References

No landmark historical reference specific to thyroid eye disease MRI, distinct from the modern comparative literature already cited, was identified as warranting separate citation. Category E is therefore not populated for this child page.

End of document — MRI Orbits for Thyroid Eye Disease / Graves' Orbitopathy — Child Protocol under the MRIninja Orbit / Visual Disorders master page — v1.0 — August 2026 Parent page: MRI Orbits — Generic Standard Protocol