Vascular Graft & Endograft Infection
Prosthetic vascular graft and endograft infection is life-threatening and often hard to confirm on CT alone. FDG PET/CT is the key functional test: using a focal (rather than diffuse) uptake pattern it has pooled sensitivity ≈91% and specificity ≈83%. Labelled-WBC SPECT/CT is highly specific and can be combined with FDG. The central rule is focal uptake = infection; diffuse homogeneous uptake = expected post-surgical/foreign-body reaction.
Non-infected grafts show mild diffuse FDG uptake for years because of a foreign-body reaction, so intensity alone is unreliable — the pattern matters. CT angiography defines anatomy and complications (peri-graft fluid, gas, pseudoaneurysm) but is limited early and after surgery, when FDG PET/CT and WBC imaging add most.
Clinical context & when to image
- Suspected graft/endograft infection with fever, sepsis or positive blood cultures and equivocal CT.
- Distinguishing infection from sterile peri-graft change after surgery.
- Localising the extent of infection and detecting distant septic foci.
- Guiding management (graft preservation vs explantation) and monitoring treatment.
Imaging strategy
- CT angiography is first-line for anatomy and complications (fluid, gas, pseudoaneurysm).
- FDG PET/CT is the principal functional test — interpret on uptake pattern, not intensity alone.
- Labelled-WBC SPECT/CT is highly specific and useful where FDG is equivocal, or combined with FDG.
- Use the PET/CT fusion to localise metabolic findings to the graft.
Protocol
- Standard whole-body FDG PET/CT with glycaemic control; delayed imaging can help.
- Fuse PET with diagnostic or CT-angiographic images where available.
- WBC SPECT/CT over the graft when performed.
- Record time since surgery — it strongly affects interpretation.
How to read it
- Infection: focal or heterogeneous intense uptake at part of the graft, often with irregular margins and peri-graft soft-tissue/fluid uptake.
- Non-infected graft: mild, diffuse, homogeneous uptake along the graft.
- Gas, fluid collections and pseudoaneurysm on CT support infection.
- Focal WBC accumulation at the graft is specific for infection.
Diagnostic performance
A meta-analysis (9 studies) found FDG PET/CT interpreted by focal uptake gave pooled sensitivity 90.6% and specificity 82.8% — the best-performing criterion. SUVmax-based criteria were similarly sensitive but less specific. A 2020 meta-analysis (13 studies) likewise found the uptake-pattern criterion most accurate (sensitivity 94%, specificity 81%).
| Method | Sensitivity | Specificity | Role |
|---|---|---|---|
| FDG PET/CT — focal pattern | ≈91% | ≈83% | Preferred functional criterion |
| FDG PET/CT — SUVmax | ≈93% | ≈70% | Sensitive; less specific |
| Labelled-WBC SPECT/CT | High | High | Specific problem-solver |
| CT angiography | Variable | Variable | First-line anatomy/complications |
Pitfalls
- Early post-operative inflammation causes false-positive FDG uptake — FDG PET/CT preferably at least 4 months after surgery; WBC imaging can be performed at any time after surgery.
- Mild diffuse uptake is expected in non-infected grafts — do not over-call intensity.
- Hyperglycaemia degrades FDG images.
- Small low-grade infections may be missed on CT alone.
Evidence & guidelines
- Meta-analysis: FDG PET/CT focal-pattern pooled sensitivity ≈91%, specificity ≈83%.
- The MAGIC (Management of Aortic Graft Infection Collaboration) criteria list suspicious FDG PET/CT activity and radiolabelled-leucocyte uptake among the minor radiological criteria for aortic graft infection.
- The EANM vascular graft infection guideline (2022) and EANM/SNMMI FDG v2.0 recommend FDG PET/CT or WBC scintigraphy when CTA is negative or equivocal and suspicion persists; FDG preferably at least 4 months after surgery, whereas WBC imaging can be performed at any time.
In depth
- The MAGIC criteria (2016) diagnose graft infection when one major criterion plus any other major or minor criterion from a different category (clinical/surgical, radiological, laboratory) is present.
- In the prospective VASGRA cohort (257 patients, 137 with graft infection), MAGIC criteria had 99% sensitivity but only 61% specificity when 'suspected' cases were counted as infected; counting only 'diagnosed' cases as infected gave 93% sensitivity and 93% specificity.
- CT angiography is the first-line test but only moderately accurate (pooled sensitivity 67%, specificity 63% in a 2018 meta-analysis); CT sensitivity is about 94% (specificity 85%) in advanced infection but only about 55% in low-grade infection.
- Across meta-analyses FDG PET/CT has high sensitivity (about 89–98%) but lower, variable specificity (about 59–81%), so a negative scan makes graft infection unlikely (very high negative predictive value), especially if performed before antibiotics.
- Interpretation methods (Rojoa 2019 meta-analysis): visual grading 89%/61%, focal-uptake pattern 93%/78%, SUVmax 98%/80%, target-to-background ratio 57%/76%; EANM regards intense, focal or heterogeneous uptake as the most reliable sign of infection, whereas diffuse homogeneous uptake is not.
- Published SUVmax cut-offs differ between studies (for example ≥3.8 and ≥4.9 in two Zurich cohorts) and no validated threshold exists; SUV is also affected by scanner, uptake time, reconstruction and blood glucose.
- Combining FDG PET with contrast-enhanced CT improves accuracy over PET/CT with low-dose CT or CT alone; myocardial suppression preparation matters mainly when the aortic root, arch or a valved conduit is being assessed.
- Non-infected synthetic grafts commonly show FDG uptake (92% in Keidar’s study), reflecting a chronic foreign-body reaction; Dacron tends to be inhomogeneous and Gore-Tex homogeneous, and uptake did not decline over up to 16 years of follow-up.
- False positives include post-operative sterile inflammation (risk is highest in the first 6–8 weeks and FDG PET is best delayed until at least 4 months after surgery), haematoma or lymphocele, and surgical adhesives such as BioGlue.
- WBC scintigraphy can be performed at any time after surgery and is favoured in the first 4 months, when FDG false positives are common; infection is shown by uptake that increases in intensity or extent over time, and for abdominal grafts early (30 min) and 2–3 h images avoid non-specific bowel activity of ⁹⁹ᵐTc-HMPAO.
Sources: PMID 27771318 (Lyons, EJVES 2016) · PMID 34140225 (Anagnostopoulos, EJVES 2021) · PMID 30122333 (Reinders Folmer, EJVES 2018) · EANM 2022 VGEI guideline (PMID 35376992) · Orton, Radiographics 2000 (PMID 10903688) · PMID 30122333 · PMID 30241981 · PMID 32645420