Infection & Inflammation Imaging: Which Test When
Choosing among labeled-leukocyte scintigraphy, FDG-PET/CT, ⁶⁷Ga, and bone/marrow imaging by clinical question
Quick Answer
There is no universal "infection scan" — the right test is dictated by the anatomic site and the clinical question. Labeled-leukocyte (WBC) scintigraphy, usually with a complementary bone-marrow (sulfur colloid) scan, is the workhorse for peripheral bone and prosthetic-joint infection, where neutrophil-mediated infection is expected. FDG-PET/CT is superior for the spine, large-vessel and metabolic/granulomatous processes, fever of unknown origin, and soft-tissue/whole-body surveys, giving high sensitivity and one-stop anatomic localization. ⁶⁷Ga persists mainly for spinal infection and some chronic/granulomatous or immunocompromised settings where it retains a niche. Matching the modality to where the infection is — and to whether you need to separate infection from marrow or from healing bone — is the whole game.
The decision in one place
Pick the test by site and question: labeled WBC ± marrow subtraction for the appendicular skeleton and prosthetic joints (it separates infection from red-marrow expansion and neuropathic remodeling); FDG-PET/CT for the spine, vessels, FUO, sarcoidosis, and whole-body soft-tissue surveys; ⁶⁷Ga as a spine/granulomatous fallback. FDG's high negative predictive value makes it a strong rule-out.
Head-to-head
| Modality | Mechanism | Best for | Weakness |
|---|---|---|---|
| Labeled leukocytes (¹¹¹In / ⁹⁹ᵐTc-HMPAO WBC) | Neutrophil migration to pyogenic infection | Prosthetic joint / peripheral osteomyelitis, diabetic foot; specific for neutrophilic infection | Labor-intensive labeling; poor in spine; misleads without a paired marrow scan |
| WBC + ⁹⁹ᵐTc-sulfur-colloid marrow | Infection localizes WBC but not marrow → spatial mismatch | Distinguishing infection from altered marrow (prostheses, Charcot foot) | Two studies; time and dosimetry |
| FDG-PET/CT | Glucose uptake by activated leukocytes | Spine, vasculitis/graft, FUO, sarcoid, soft tissue; fast, whole-body, high NPV | Nonspecific (post-op, tumor, healing); glucose control needed |
| ⁶⁷Ga-citrate | Transferrin/lactoferrin binding | Spinal infection, chronic/granulomatous, some immunocompromised | Poor resolution, delayed multiday imaging, high dose |
How to sequence
By scenario:
- Prosthetic joint infection / appendicular osteomyelitis: labeled WBC scan with a paired marrow (sulfur colloid) scan — infection produces WBC activity that does not match marrow; congruent activity is marrow, not infection.
- Diabetic foot / Charcot (neuropathic) joint: WBC + marrow subtraction separates osteomyelitis from neuropathic remodeling; MRI is complementary.
- Vertebral osteomyelitis / discitis: FDG-PET/CT (or ⁶⁷Ga) — labeled WBC is unreliable in the spine (photopenic or nonspecific marrow uptake).
- Fever of unknown origin, vascular graft/large-vessel vasculitis, sarcoidosis, whole-body survey: FDG-PET/CT for sensitivity and one-stop localization.
- Endocarditis / cardiac device infection: FDG-PET/CT (with proper cardiac suppression prep) plus, for prosthetic valves, correlation with CT.
Why the marrow scan matters
Labeled leukocytes accumulate in both infection and hematopoietically active marrow. In regions of marrow expansion or altered distribution (around prostheses, in the neuropathic foot), WBC uptake alone is ambiguous. Adding a sulfur-colloid marrow scan resolves it: infection = WBC activity that is spatially incongruent with (exceeds) marrow activity, whereas congruent activity is simply marrow. This subtraction logic is the single most important selection principle in skeletal infection imaging.
Pitfalls in selection
- FDG after surgery: post-operative inflammation is FDG-avid for weeks to months; timing and pattern matter.
- Glucose control is required for reliable FDG; hyperglycemia degrades sensitivity.
- Spine + labeled WBC = trap: vertebral infection is often photopenic on WBC — do not read a "negative" WBC spine study as excluding infection; use FDG/⁶⁷Ga.
Bottom line
Choose by site: WBC ± marrow for peripheral bone and prostheses, FDG-PET/CT for spine, vessels, FUO, and soft-tissue/whole-body, and ⁶⁷Ga as a spinal/granulomatous fallback.
Self-Check
Q1. Which infections are best imaged with labeled leukocytes, and why is a marrow scan usually added?
Answer: Prosthetic-joint and peripheral (appendicular) osteomyelitis. A paired sulfur-colloid marrow scan is added because WBCs localize to both infection and marrow; infection is WBC activity incongruent with marrow, distinguishing it from marrow expansion/remodeling.
Q2. Why is FDG-PET/CT preferred over labeled WBC for vertebral osteomyelitis?
Answer: Labeled WBC is unreliable in the spine (spinal infection is often photopenic or shows nonspecific marrow uptake); FDG-PET/CT (or ⁶⁷Ga) is far more sensitive there.
Q3. For which clinical problems is FDG-PET/CT the modality of choice?
Answer: Spine infection, large-vessel/graft vasculitis, fever of unknown origin, sarcoidosis, and whole-body soft-tissue surveys — high sensitivity, whole-body coverage, and strong negative predictive value.
Q4. What is the interpretive rule that defines infection on a combined WBC/marrow study?
Answer: Spatial incongruence — WBC uptake that exceeds/does not match the marrow distribution indicates infection; congruent WBC and marrow activity is normal marrow, not infection.
Key References
- SNMMI/EANM joint guidelines on labeled-leukocyte scintigraphy and ¹⁸F-FDG imaging of infection/inflammation.
- Society guidance on prosthetic-joint infection, vertebral osteomyelitis, vascular graft infection, and FUO imaging.
Evidence & sources
https://nuc.med/infection-imaging-selection/ Report an issue