1. Executive Summary

Whole-body MRI (WB-MRI) is now the recommended first-line imaging modality for multiple myeloma (MM) staging and response assessment in current international guidelines, having replaced whole-body CT (low-dose skeletal survey CT) and bone scintigraphy as the standard of care in myeloma bone disease evaluation [1, 2, 3]. This paradigm shift — completed by the incorporation of WB-MRI into the International Myeloma Working Group (IMWG) 2014 criteria [1] and reinforced by the British Society of Haematology (BSH) 2019 guidelines [2] — reflects a fundamental advantage: WB-MRI detects bone marrow infiltration by malignant plasma cells before bone destruction occurs, providing earlier and more accurate staging information than any skeletal imaging modality that relies on cortical disruption for detection.

The clinical impact is substantial: up to 30–40% of patients classified as non-secretory or early-stage by laboratory and CT criteria are upstaged when WB-MRI is applied, revealing previously undetected focal lesions and diffuse marrow infiltration patterns that change both prognosis and treatment decision [4, 5]. For patients with smouldering myeloma (SMM), WB-MRI has become the pivotal modality for risk stratification: the presence of ≥ 2 unequivocal focal lesions on WB-MRI in SMM constitutes a myeloma-defining event (MDE) that mandates treatment initiation regardless of laboratory parameters [1].

1.1 Core Strengths

Pre-erosive bone marrow detection: MRI detects malignant plasma cell infiltration as T1-hypointense, STIR-hyperintense signal replacing the normally T1-bright yellow fatty marrow. This signal change precedes cortical bone erosion or destruction by weeks to months, enabling staging at a biologically earlier disease stage than CT or plain radiography.

Diffuse vs. focal pattern discrimination: WB-MRI uniquely differentiates the four myeloma infiltration patterns — focal (discrete lesions), diffuse (homogeneous marrow replacement), combined (focal on diffuse), and variegated (patchy inhomogeneous) — which have distinct prognostic significance and treatment response profiles [6].

Spinal cord compression detection: WB-MRI is the only whole-body staging modality that simultaneously assesses cord compression risk from vertebral lesions, identifies epidural extension, and detects extramedullary disease without requiring a separate dedicated spine MRI.

Soft tissue plasmacytoma and extramedullary disease: extramedullary plasmacytoma (EMP) — which occurs in approximately 10–15% of newly diagnosed MM and carries a particularly poor prognosis — is detected on WB-MRI T1, STIR, and DWI sequences. CT detects EMP only when it exceeds the soft tissue attenuation threshold; PET detects metabolically active EMP; MRI detects both active and relatively quiescent EMP.

No ionising radiation: critical for a disease requiring serial imaging across a multi-year treatment trajectory. WB-MRI replaces the cumulative radiation exposure of repeated low-dose CT skeletal surveys.

Response assessment superiority over CT: CT cannot detect bone marrow response — marrow reconstitution after successful treatment appears as T1-bright fatty marrow recovery on MRI, which is invisible on CT. WB-MRI response criteria (MRI response criteria from IMWG [1]) track marrow signal changes that correlate with depth of haematological response.

1.2 Intrinsic Limitations of the Generic Protocol

Acquisition time: a complete WB-MRI with DWI requires 45–70 minutes depending on the number of stations, DWI parameters, and optional sequences included. This is substantially longer than whole-body CT (5–10 minutes) and represents a significant scanner time commitment. Patient comfort, breath-hold compliance, and claustrophobia are practical barriers.

Active hyperaemia and red marrow reconversion false positives: physiological diffuse T1 hypointensity and STIR hyperintensity from red marrow reconversion (in anaemia, cytokine administration, or haematopoietic stimulant treatment) can simulate diffuse myeloma infiltration on STIR-only protocols. DWI and T1 assessment in combination are necessary to distinguish physiological from pathological marrow signal.

Cortical bone and fracture detail inferior to CT: vertebral compression fractures are visible on WB-MRI, but the assessment of cortical integrity, trabecular architecture, and subtle non-vertebral fractures is less detailed than CT. For peri-operative surgical planning of impending pathological fractures, CT or cone-beam CT complements WB-MRI.

ADC values post-treatment: ADC in treated myeloma lesions may transiently increase (oedema, necrosis) before true marrow recovery, complicating early response assessment. The interpretation window for ADC-based response assessment is approximately 3 months after treatment.

PET/CT complementarity: FDG-PET/CT and WB-MRI are complementary rather than equivalent. PET detects metabolically active disease; MRI detects bone marrow infiltration regardless of metabolic activity. Non-secretory myeloma, treated lesions with reduced metabolism, and small extramedullary deposits may be better detected on MRI than PET. Conversely, metabolically active bone lesions may be PET-positive before they become MRI-visible.

When dedicated child protocols are required: SMM with WB-MRI for MDE assessment (requires specific focal lesion counting methodology); post-transplant response assessment (specific ADC threshold application); spinal cord compression (dedicated spine protocol supplementing WB-MRI); plasmacytoma (local staging protocol); paediatric plasma cell dyscrasias; AL amyloidosis with cardiac involvement (cardiac MRI supplement).

2. Clinical Indications and Role of WB-MRI in Myeloma

2.1 Standard Indications

Initial staging of newly diagnosed MM is the primary indication. WB-MRI replaces the skeletal survey at this timepoint in centres that have implemented IMWG 2014 guidelines [1]. The staging WB-MRI defines: the number and distribution of focal lesions; the pattern of marrow infiltration; the presence of diffuse marrow replacement; spinal cord compression risk; extramedullary disease. These findings influence treatment stratification and are recorded as the baseline for all subsequent response comparisons.

Smouldering myeloma (SMM) risk stratification has become one of the most impactful WB-MRI applications. IMWG 2014 criteria [1] define ≥ 2 focal lesions on WB-MRI (each ≥ 5 mm) in a patient meeting SMM criteria as a myeloma-defining event, equivalent to end-organ damage in terms of treatment requirement. This MRI-based MDE has changed clinical management for a substantial proportion of patients previously managed with watchful waiting.

Treatment response assessment at mid-treatment and post-treatment timepoints. WB-MRI response categories (complete response = disappearance of all lesions with T1 normalisation; partial response = reduction in number and STIR signal; stable disease; progressive disease) are defined by IMWG criteria and have been validated against haematological response endpoints [1, 6].

Relapse assessment when laboratory parameters suggest biochemical relapse and CT or PET findings are non-contributory. WB-MRI identifies new focal lesions, diffuse marrow re-infiltration, or new extramedullary disease earlier than skeletal survey or CT.

Spinal cord compression risk stratification before surgical or radiotherapy planning. WB-MRI identifies vertebral lesions with epidural extension, cord displacement, or cord signal change — information that requires immediate clinical action.

2.2 Role in the Myeloma Staging and Treatment Decision Framework

WB-MRI operates within the IMWG 2014 diagnostic framework [1] where it specifically contributes to:

2.3 Urgent Red Flags

3. Preparation Reference

Universal MRI safety screening belongs to the general MRI preparation page and is not repeated here.

3.1 Anatomy-Specific Preparation Items

Analgesic management: myeloma patients frequently have significant bone pain — particularly vertebral, pelvic, and long bone pain. The extended scanning duration (45–70 minutes) may be intolerable without adequate pre-examination analgesia. Co-ordinate with the haematology team: oral analgesia (paracetamol, low-dose opioid) administered 30–60 minutes before the examination significantly improves patient compliance and reduces motion artefacts from pain-related movement.

No specific dietary preparation is required for WB-MRI in myeloma (unlike hepatic MRI or biliary MRCP). The patient may eat and drink normally before the examination.

Gadolinium: WB-MRI for myeloma is performed without gadolinium contrast for the standard protocol. Post-contrast T1 may be conditionally added for extramedullary plasmacytoma characterisation or soft tissue mass assessment, but the core diagnostic sequences (T1, STIR, DWI) are non-contrast. Renal function assessment is required only when contrast is planned.

Implants relevant to WB-MRI: myeloma patients may have orthopaedic fixation hardware (intramedullary nails for pathological fracture prevention, vertebroplasty cement, kyphoplasty balloons). Metal implants produce susceptibility artefacts that limit assessment of adjacent bone marrow on WB-MRI. Document all prior orthopaedic interventions. Cement from vertebroplasty/kyphoplasty appears T1-dark and STIR-dark within the vertebral body — this must not be misinterpreted as a persistent lesion.

Granulocyte colony-stimulating factor (G-CSF): patients who have recently received G-CSF, pegfilgrastim, or plerixafor for stem cell mobilisation will have generalised diffuse bone marrow hyperaemia and red marrow reconversion on WB-MRI, simulating diffuse myeloma infiltration. If possible, WB-MRI should be deferred until at least 3 weeks after G-CSF administration. If deferral is not clinically possible, document the G-CSF timing prominently in the report and interpret the diffuse marrow signal with appropriate caution.

3.2 Patient Positioning on the MRI System

Position: supine, head-first, arms alongside the body. Arms alongside the body (not above the head) is standard for WB-MRI because: (a) arm-up positioning is uncomfortable for 45–70 minutes; (b) arms alongside does not significantly compromise FOV coverage for the axial skeleton, which is the primary target.

Number of stations and coverage: the WB-MRI requires multiple consecutive acquisitions (stations) to cover the entire skeleton. Standard WB-MRI in myeloma covers:

• Station 1: skull / brain (optional) and cervical spine
• Station 2: thorax including thoracic spine, ribs, and sternum
• Station 3: abdomen including lumbar spine and pelvis
• Station 4: proximal femora bilaterally (to the knee in some protocols)
• Station 5: lower legs (seldom required for myeloma — cortical disease rare)

The minimum oncological WB-MRI for myeloma covers the axial skeleton from skull vertex to proximal femora — typically achieved in 4 stations.

Coil selection: phased-array surface coils are used at each station, repositioned between stations. At each station, the combination of anterior body matrix coil + posterior spine coil elements provides the multi-channel reception needed for parallel imaging acceleration. Some modern scanners (Siemens MAGNETOM Vida, GE Signa, Philips Elition) offer table-mounted rolling coil platforms that reduce repositioning time between stations.

Centring: each station must be centred for that anatomical region. The lumbar spine and pelvis station is the most critical — centre on L3 to ensure the full lumbar spine and iliac crests are included.

Patient comfort: foam pads under the knees reduce lumbar lordosis and improve patient comfort for the full examination duration. A thermal blanket improves thermal comfort and reduces shivering motion. Earplugs and communication with the patient between stations are essential for a 60-minute examination.

4. Standard Protocol Design

4.1 Mandatory Core Sequences

The IMWG-endorsed WB-MRI protocol [1, 2] and the British Society of Haematology guidelines [2] specify the following minimum sequences:

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

Coronal T1 (non-fat-suppressed) is the foundational sequence for myeloma WB-MRI. The rationale is specific to myeloma bone marrow biology: normal adult yellow marrow (predominantly present in the axial skeleton after age 25) is T1-bright due to fat content. Malignant plasma cell infiltration replaces this fat with cellular tumour tissue → T1-hypointense signal. The T1 sequence provides: (a) direct visual comparison of normal yellow marrow vs. infiltrated marrow; (b) lesion detection as discrete T1-dark foci within T1-bright background; (c) vertebral body morphology; (d) soft tissue assessment. Fat suppression must NOT be applied to the T1 sequence used for marrow assessment — fat suppression eliminates the T1-bright background that makes T1-dark lesions conspicuous.

Key T1 appearances by pattern:

• Focal: discrete T1-dark lesions within T1-bright background — the most common pattern in early myeloma
• Diffuse: uniform T1-hypointensity replacing the entire marrow — indicates heavy diffuse infiltration (> 50% plasma cell burden usually)
• Variegated/inhomogeneous: patchy intermediate signal — common in treated or partially responded marrow
• Normal: uniform T1-bright marrow — does not exclude disease (MRI-negative marrow biopsy can still show positive)

STIR is the T2-weighted fat-suppressed complement to the coronal T1. On STIR, malignant marrow infiltration appears STIR-hyperintense (long T1+T2 of plasma cells) against the nulled fat background. STIR is particularly sensitive for: (a) bone marrow oedema adjacent to lesions; (b) diffuse marrow signal changes that may be subtle on T1; (c) soft tissue extramedullary extension from vertebral lesions; (d) perilesional inflammatory change; (e) cortical disruption with adjacent soft tissue signal. The STIR must always precede any gadolinium injection — post-gadolinium STIR is contraindicated (see STIR sequence page for the physical basis of this absolute contraindication across all MRIninja protocols).

DWI + ADC map has transformed myeloma WB-MRI from a morphological examination into a functional assessment tool. The DWI signal in myeloma reflects tumour cellularity: densely packed plasma cells restrict water diffusion → high DWI signal, low ADC. Published thresholds and clinical applications:

• Active focal lesions: ADC typically 0.5–1.0 × 10⁻³ mm²/s (below normal yellow marrow)
• Normal yellow marrow: ADC 0.2–0.4 × 10⁻³ mm²/s at low b-value; higher at b=900
• Treated, responding lesions: ADC increases as cellularity decreases and fat reconversion occurs
• Red marrow (physiological): ADC intermediate (0.4–0.8 × 10⁻³ mm²/s) — overlaps with low-grade myeloma; clinical context essential

The b-value selection determines the balance between signal and suppression of background tissue. The standard MRIninja WB-MRI for myeloma uses b=50 (low b, provides high SNR reference and vascular suppression) and b=900 (diagnostic b, sufficient for marrow lesion detection while maintaining adequate SNR at WB coverage with parallel imaging).

b=0 vs b=50: using b=50 rather than b=0 as the low b-value suppresses vascular signal (blood moves during the diffusion gradient interval), reducing the conspicuity of vascular structures that might otherwise simulate marrow lesions on low-b images.

Sagittal T2 spine (strongly recommended): the sagittal T2 TSE of the thoracic and lumbar spine is essential for: (a) vertebral fracture assessment; (b) endplate and cortical integrity; (c) spinal cord signal and position relative to lesions; (d) epidural disease extension. This sequence is acquired in the sagittal plane and cannot be replaced by the coronal T1/STIR. The thoracic spine is particularly difficult to assess on coronal views due to rib overlay.

4.4 Sequence Matching and Cross-Sequence Consistency

For serial response assessment, geometric consistency is paramount. The coronal T1 and STIR must use identical slice position, thickness, and FOV at each follow-up. If even one station is slightly shifted, the same vertebral body may appear in different positions, making direct comparison of signal intensity and lesion dimensions unreliable.

Practical implementation: document the first examination's station positions (centre coordinates, slice angulation) in the RIS or PACS and reproduce these at each follow-up. Most WB-MRI platforms allow storage of geometric parameters as a "preferred exam" or "protocol template" for reproducibility.

The DWI must also use consistent b-values and geometric parameters across examinations. ADC changes at follow-up are only interpretable if the acquisition parameters were constant.

4.5 Fat Suppression in WB-MRI for Myeloma

STIR is the mandatory fat suppression technique for WB-MRI in myeloma (not Dixon, not SPAIR). The reasons are specific to this clinical context:

1. Off-isocentre reliability: WB-MRI requires multiple stations, each positioned at a different location along the table. At the thoracic station, the imaging volume may be 15–20 cm from isocentre. Spectral fat suppression (SPAIR, CHESS) fails predictably off-isocentre due to B0 inhomogeneity from the air-tissue interfaces at the lungs. STIR provides B0-independent fat suppression at every station, including the thorax.

1. Additive T1+T2 contrast in STIR: as documented in the STIR sequence page, STIR provides additive T1+T2 contrast that makes marrow oedema and tumour infiltration more conspicuous than T2-FS alone. This additive effect is particularly valuable in myeloma where marrow infiltration produces concurrent T1 and T2 changes.

1. STIR TI calibration per station field strength: the TI for STIR must be correct for the scanner field strength (150–175 ms at 1.5T; 200–230 ms at 3T). This does not change between stations but must be verified during protocol commissioning.

Dixon for fat quantification: Dixon-based fat fraction mapping (see Pancreas master page and Liver master page for Dixon principles) is a conditional addition to the WB-MRI for myeloma research protocols. Fat fraction maps distinguish red marrow (low fat fraction, 30–50%) from yellow marrow (high fat fraction, 70–90%) from plasma cell infiltration (very low fat fraction, < 20%). Dixon fat quantification is not required for clinical staging but is used in clinical trials assessing depth of response.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

Motion artefacts — primary quality failure in WB-MRI: the long acquisition time (45–70 minutes) increases the probability of patient motion between stations. This is particularly relevant for the DWI stations, which are the most motion-sensitive. Mitigation: (a) adequate pre-examination analgesia; (b) patient communication between each station; (c) head cushioning to prevent unconscious head movement; (d) shorter acquisition per station with more stations (reduces per-station breath-hold demand).

EPI geometric distortion in DWI: the EPI readout used for DWI is sensitive to B0 inhomogeneity from air-tissue interfaces — particularly at the thorax (lung-mediastinum interface) and the sacrum (bowel gas-bone interface). Geometric distortion produces apparent displacement of vertebral body signal relative to the true anatomical position. At 3T, distortion is approximately 2× that at 1.5T. Mitigation: (a) 1.5T has substantially less EPI distortion than 3T for WB-DWI; (b) parallel imaging (iPAT/SENSE) reduces echo train length and geometric distortion; (c) reduced FOV techniques (FOCUS/iZOOM/ZOOMit) further reduce distortion at specific stations; (d) B0-corrected EPI available on Siemens and GE platforms; (e) accept some distortion and correlate DWI with co-registered T1/STIR for anatomical localisation.

STIR fat suppression failure in the thorax: the thorax is the most challenging region for fat suppression in WB-MRI due to the large B0 disturbance from the air-lung interface. STIR provides the most reliable fat suppression at this station (B0-independent) — confirming the choice of STIR over SPAIR for all WB-MRI stations. Even with STIR, residual incomplete fat suppression in the posterior ribs and paravertebral regions may occur; this should be noted in the report when it affects interpretation.

Cardiac motion artefact at the thoracic station: cardiac motion produces pulsation artefacts in the A-P direction (for axial DWI) and ghosting in the phase direction (for coronal STIR). Mitigation: cardiac gating (rarely practical for WB-MRI due to time constraints); triggering the DWI acquisition to the cardiac cycle using a peripheral pulse trigger (available on all major platforms) substantially reduces cardiac artefact at the thoracic station at the cost of approximately 30% longer acquisition time; alternatively, ECG-triggered acquisition. For coronal STIR at the thoracic station, saturation bands superior and inferior to the heart reduce propagation of cardiac ghost artefacts.

Background tissue signal on DWI MIP (DWIBS): the DWIBS (diffusion-weighted whole-body imaging with background body signal suppression) technique generates MIP projections of the inverted DWI at high b-value, creating a PET-equivalent "dark body, bright lesion" display. At b=900, some normal structures (spinal cord, kidneys, lymph nodes, ganglia) have high DWI signal and appear as "lesions" on DWIBS MIP. Always interpret DWIBS MIP in combination with the source DWI images and the ADC map — do not diagnose lesions on DWIBS MIP alone.

5.2 Protocol Efficiency and Throughput

A complete BSH/IMWG-compliant WB-MRI for myeloma — coronal T1 + coronal STIR + axial DWI (3–4 stations each) + sagittal T2 spine — requires:

Compressed sensing and parallel imaging (iPAT factor 2–3) can reduce per-station acquisition time by 30–50%, achieving a complete WB-MRI in 35–45 minutes at 3T.

Abbreviated WB-MRI for SMM monitoring (consensus expert practice, not guideline-mandated): some centres use a shorter protocol for serial SMM monitoring — coronal T1 + coronal STIR only (no DWI) — in approximately 30 minutes, primarily aimed at lesion counting for MDE surveillance. This abbreviated approach is not adequate for response assessment in treated myeloma where DWI is required for ADC-based evaluation.

5.3 Field Strength Considerations

1.5T is currently the preferred field strength for clinical WB-MRI in myeloma at most expert centres, for the following reasons:

EPI geometric distortion: at 1.5T, EPI distortion in the DWI is substantially less than at 3T, producing more anatomically accurate DWI images particularly at the thoracic station adjacent to the lungs. This is the primary reason myeloma WB-MRI guidelines do not mandate 3T.

STIR performance: STIR is equally effective at 1.5T and 3T (with appropriate TI adjustment). The advantage of 3T for spectral fat suppression (SPAIR) is irrelevant since STIR is used.

SAR at 3T for STIR: STIR at 3T with 3–4 stations produces significant SAR accumulation. The extra 180° inversion pulse per TR at 3T may trigger automatic TR extension, which shifts the STIR TI-null point (see STIR sequence page). Protocol monitoring for automatic TR changes is essential at 3T.

3T advantages: higher SNR enabling thinner DWI slices and faster acquisition; improved spatial resolution for focal lesion characterisation; 3T may become preferred as EPI correction algorithms improve.

6. Contrast Use Principles Specific to WB-MRI in Myeloma

6.1 Non-Contrast Standard Protocol — Sufficient For

The standard WB-MRI protocol for myeloma staging, response assessment, and SMM surveillance is performed without gadolinium contrast. Non-contrast WB-MRI (T1, STIR, DWI) is diagnostically adequate for:

• Initial staging of newly diagnosed MM (focal lesion detection, infiltration pattern)
• SMM MDE assessment (focal lesion counting per IMWG criteria)
• Response assessment (T1 signal recovery, STIR normalisation, ADC changes)
• Relapse detection (new focal lesions, diffuse re-infiltration)
• Spinal cord compression risk assessment (combined with sagittal T2 spine)

6.2 Gadolinium Indicated — Region-Specific Contexts

Post-contrast T1 sequences are conditionally required for:

• Extramedullary plasmacytoma characterisation: EMP enhances intensely and uniformly; enhancement pattern distinguishes EMP from lymph node or reactive soft tissue mass
• Epidural disease extent: post-contrast T1 sagittal spine delineates the enhancing epidural tumour margin and its relationship to the cord
• Leptomeningeal involvement: post-contrast T1 axial brain demonstrates leptomeningeal enhancement in the rare event of CNS myeloma
• Post-treatment assessment of ambiguous lesions: persistent T1-hypointense foci with uncertain biological activity may be better characterised with enhancement assessment

When gadolinium is administered, it must be given after all STIR acquisitions — gadolinium shortens T1 and would compromise STIR fat nulling (see STIR sequence page absolute contraindication).

6.3 Post-Contrast Timing

Standard post-contrast T1 fat-suppressed (Dixon or SPAIR) at 3–5 minutes post-injection provides equilibrium phase enhancement characterisation of EMP and epidural disease. No specific arterial phase or delayed phase acquisition is routinely required for myeloma applications.

7. Reporting Essentials

7.1 Interpretation Framework: The MY-RADS Classification

The MY-RADS (Myeloma Response Assessment and Diagnosis System) framework [6] — developed by the International Myeloma Working Group and published collaboratively — provides the standardised reporting system for WB-MRI in myeloma. Radiologists reporting myeloma WB-MRI should be familiar with the MY-RADS categories:

MY-RADS 1: definitely not myeloma (normal marrow; benign lesion)

MY-RADS 2: probably not myeloma (Schmorl nodes, haemangioma, simple cysts — benign features)

MY-RADS 3: equivocal for myeloma (cannot distinguish between myeloma and non-myeloma cause)

MY-RADS 4: probably myeloma (focal T1-dark, STIR-bright lesion with restricted diffusion in appropriate clinical context)

MY-RADS 5: definitely myeloma (definitive imaging features of myeloma — e.g., cord compression from lytic vertebral lesion)

Response MY-RADS categories (for follow-up):

• MR1: complete imaging response (resolution of all lesions; T1 signal normalisation)
• MR2: partial response (lesion reduction ≥ 50%; STIR signal decrease; ADC increase)
• MR3: stable disease
• MR4: progressive disease (new lesions or ≥ 25% increase in lesion volume/number)

7.2 Mandatory Reporting Checklist

Marrow pattern (axial skeleton):

• [ ] Pattern: focal / diffuse / combined / variegated / normal
• [ ] Number of focal lesions (count those ≥ 5 mm separately for IMWG MDE)
• [ ] Largest focal lesion size (mm)
• [ ] Diffuse marrow signal: T1 / STIR / DWI / ADC characterisation
• [ ] Red marrow distribution (normal for age) vs. pathological infiltration

Spine (each vertebral level):

• [ ] Vertebral fractures: acute vs. chronic; height loss; retropulsion
• [ ] Cord compression: present / absent; level; degree; cord signal change
• [ ] Epidural disease: present / absent; level; maximum extent
• [ ] Foraminal compromise: present / absent

Extramedullary disease:

• [ ] Soft tissue plasmacytomas: location, size, DWI signal
• [ ] Pelvic and abdominal lymphadenopathy
• [ ] Pleural or pericardial effusion (if visible)

Skull and ribs:

• [ ] Focal skull lesions
• [ ] Rib lesions (note lytic from cortical disruption)
• [ ] Sternal lesion (indicates EMP risk)

Response assessment (if follow-up):

• [ ] MY-RADS response category (MR1–MR4)
• [ ] ADC change from baseline (if ADC maps available)
• [ ] New lesions: present / absent
• [ ] Change in lesion number and size

7.3 Structured Reporting

Reports must include: Indication (staging / SMM / response assessment / relapse); Technique (field strength, stations covered, b-values, whether STIR was pre- or post-contrast); Comparison (prior WB-MRI date and findings summary); Findings (marrow pattern, lesion count and size, spine assessment, EMP); MY-RADS classification (per lesion basis for equivocal lesions; overall response category for follow-up); IMWG MDE statement if SMM; Impression; Recommendations (biopsy target if equivocal; dedicated spine MRI if cord compromise; clinical urgency statement).

7.4 Incidental Findings — Clinical Decision Framework

Usually benign: vertebral haemangioma (T1-bright, STIR-bright, well-circumscribed, does not restrict diffusion); Schmorl nodes (endplate depression with sclerotic margin); simple renal or hepatic cysts; gallstones.

May require follow-up: moderate lymphadenopathy (< 10 mm short axis) without restriction; bone lesion with MY-RADS 3 features in a low-suspicion clinical context; incidental focal bone lesion not in keeping with myeloma (consider fibrous dysplasia, Paget disease — have specific MRI patterns).

Urgent communication required: cord compression with signal change; new pathological fracture; unexpected organ invasion from plasmacytoma; unsuspected pleural disease; unexpected abdominal mass.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

1. WB Scout ← essential first step; defines all station positions for the entire examination
2. Sagittal T2 spine (thoracic + lumbar) ← acquire early; critical for cord compression detection; if patient deteriorates and exam must be stopped, cord assessment is complete
3. Coronal T1 (all stations) ← primary marrow assessment; next most time-efficient order
4. Coronal STIR (all stations) ← oedema and infiltration; always pre-contrast
5. Axial DWI (all stations) ← functional assessment; longest per-station time if cardiac triggered

8.2 Positioning Tricks

For patients with kyphosis: a wedge foam pad under the thoracic region improves contact with the posterior spine coil and reduces the gap between the patient's thoracic spine and the table surface, improving SNR at the thoracic station.

For patients with hip prostheses: the metal-induced susceptibility signal loss from hip prostheses will affect DWI and STIR at the pelvic station. Use STIR (not SPAIR) and document the limitation. At 1.5T, the signal loss radius is approximately 2–5 cm from the prosthesis; at 3T, it is 4–10 cm.

For patients with severe pain limiting supine positioning: elevate the head of the table to 30–45° (reverse Trendelenburg or pillow elevation) to allow partial sitting position — acceptable compromise for the thoracic and cervical stations.

8.3 Fast Salvage Protocol

Total: approximately 48 minutes — covers the critical diagnostic requirements. DWI at only 2 stations (spine + pelvis, the highest-yield regions) reduces time while retaining most functional information.

8.4 Common Avoidable Errors

9. Quality Control Checklist

• [ ] WB scout coverage from skull vertex to proximal femora
• [ ] All stations overlapping by ≥ 3 cm at junctions
• [ ] Coronal T1 acquired without fat suppression
• [ ] STIR acquired before any gadolinium injection
• [ ] G-CSF history documented and noted in technique
• [ ] DWI b-values documented (b=50 + b=900 or equivalent per protocol)
• [ ] ADC map generated and available for review
• [ ] Sagittal T2 spine (thoracic + lumbar) completed
• [ ] Cord compression assessment completed on sagittal T2
• [ ] STIR motion assessment: no station with unacceptable motion
• [ ] DWI distortion at thoracic station acceptable for diagnostic use
• [ ] DWIBS MIP generated from DWI (if used for reporting)
• [ ] All stations correctly labelled (station number, anatomical level)
• [ ] Serial exam: station positions reproduced from baseline

11. Evidence Gaps and Ongoing Debate

WB-MRI vs PET/CT at initial staging: the IMWG 2014 criteria [1] accept both WB-MRI and PET/CT as appropriate staging modalities. No prospective randomised trial has directly compared their impact on treatment decisions and outcomes in newly diagnosed MM. The choice is currently institution-dependent, with some expert centres using both in parallel, particularly for high-risk presentations.

Optimal DWI b-values: published WB-DWI protocols for myeloma use b-values ranging from b=50–100 (low) and b=800–1000 (diagnostic). No prospective study has formally compared diagnostic accuracy across b-value combinations for myeloma focal lesion detection. The b=50 and b=900 combination is the current consensus [2] but is not based on formal optimisation data.

ADC response assessment thresholds: the ADC increase associated with treatment response has been described in multiple series, but validated thresholds for "complete ADC response" remain centre-dependent and not universally standardised. IMWG response criteria [1] do not specify a minimum ADC change threshold for radiological complete response.

Abbreviated WB-MRI for SMM surveillance: whether coronal T1 + STIR alone (without DWI) is adequate for SMM focal lesion counting for MDE assessment has not been formally validated against the full protocol. The 2-lesion threshold for MDE was established on protocols that included DWI in most cases; pure morphological assessment may have different sensitivity.

3T for WB-MRI: despite theoretical SNR advantages, 3T for clinical WB-DWI in myeloma has not been shown to improve diagnostic accuracy or change clinical outcomes compared with 1.5T in prospective comparative studies. The EPI distortion disadvantage at 3T remains the primary argument for 1.5T.

AI-assisted lesion counting: automated focal lesion detection and counting on WB-MRI is under active development. Preliminary results show sensitivity 75–90% and specificity 80–92% for focal lesion detection at b=900. No clinically validated, regulatory-cleared tool exists at the time of writing.