MRI With Contrast Alone is Insufficient—You Need Both Non-Contrast and Contrast Sequences
For evaluating osteomyelitis and abscess, the American College of Radiology recommends MRI with and without IV contrast (rated 9, "usually appropriate") rather than contrast-enhanced sequences alone. 1
Why Both Sequences Are Necessary
Non-Contrast Sequences Detect the Bone Infection Itself
- Non-contrast MRI has high sensitivity (82-100%) and specificity (75-96%) for diagnosing osteomyelitis based on marrow signal changes alone. 2
- The hallmark finding of osteomyelitis—decreased T1-weighted bone marrow signal and increased signal on fluid-sensitive sequences (T2/STIR)—is visible without contrast. 3
- A normal marrow signal on non-contrast MRI reliably excludes infection with 100% negative predictive value. 3
- Fluid-sensitive sequences alone may be sufficient for diagnosis in pediatric patients, with one study showing 100% identification of septic arthritis and osteomyelitis using only these sequences. 2
Contrast Sequences Evaluate Soft-Tissue Complications
- Contrast administration significantly improves detection of abscesses, with studies showing a 46% increase in reader confidence for abscess diagnosis. 4
- Contrast helps identify the extent of soft-tissue infection, purulent collections, sinus tracts, and vascular involvement. 1, 3
- Rim-enhancing abscesses versus non-enhancing necrotic material can only be distinguished with contrast, which is critical for surgical planning. 5
- Contrast-enhanced sequences help differentiate vascularized inflammatory tissue from non-vascularized fluid collections requiring drainage. 5, 6
The ACR Rating System Clarifies This
- MRI with and without contrast receives a rating of 9 ("usually appropriate") for suspected osteomyelitis with cellulitis, wounds, or post-surgical states. 1
- MRI without contrast alone receives only a rating of 7-8, indicating it is less optimal. 1
- This rating difference reflects the superior diagnostic performance when both sequences are obtained together. 1
Clinical Algorithm
Start with plain radiographs to exclude alternative diagnoses and assess for chronic changes like periosteal reaction or sequestrum. 2, 3
Order MRI with and without IV contrast as the next imaging study when osteomyelitis or abscess remains suspected. 1
If IV contrast is contraindicated (renal insufficiency, allergy), non-contrast MRI alone is still highly accurate for detecting bone marrow changes but will have reduced sensitivity for soft-tissue abscesses. 1, 3
For surgical planning, contrast-enhanced sequences are essential to identify areas of bone devitalization, abscess formation requiring drainage, and the precise extent of infection before debridement. 3, 5
Common Pitfalls to Avoid
- Do not order contrast-only sequences—you will miss the baseline marrow signal abnormalities that establish the diagnosis of osteomyelitis itself. 2, 3
- Contrast enhancement does not improve diagnosis of peripheral osteomyelitis per se, but it dramatically improves evaluation of soft-tissue infections and abscesses. 2
- In pediatric patients, contrast increases diagnostic confidence most significantly when bone or soft-tissue edema is present on non-contrast images, but adds little value when edema is clearly absent. 4
- Early osteomyelitis (<14 days) may show only soft-tissue changes on any imaging modality, so clinical correlation remains essential. 3
Special Considerations
- For diabetic foot osteomyelitis, the International Working Group on the Diabetic Foot considers non-contrast MRI acceptable for diagnosis, but contrast is recommended for surgical planning and identifying abscesses. 3
- In children with suspected septic arthritis, decreased femoral head enhancement on early post-contrast imaging is a reliable feature distinguishing septic arthritis from transient synovitis. 2
- For hardware-associated infection, MRI with and without contrast (rating 9) helps evaluate hardware complications, though extensive metal artifact may necessitate CT with contrast (rating 6-7) or nuclear medicine studies instead. 1