Nucpaedia
Nucpaedia
GI · Bleeding

GI Bleeding Scintigraphy

Snapshot

⁹⁹ᵐTc-labelled red-cell scintigraphy detects and regionalises lower-gastrointestinal bleeding, sensitive to slow or intermittent haemorrhage (as low as ≈0.1 mL/min clinically) that catheter angiography can miss. It typically localises bleeding to a bowel region to direct angiography, endoscopy or surgery, rather than pinpointing the exact vessel.

Autologous red cells are labelled and re-injected; extravasated activity appears in the bowel lumen and moves with peristalsis. Prolonged or repeated imaging captures intermittent bleeds.

~0.1 mL/minDetection threshold
IntermittentSerial imaging helps
RegionalisesGuides next test

When to image

  • Overt mid- or lower-GI bleeding once an upper-GI source is excluded, including intermittent or obscure overt bleeding.
  • Before or to guide catheter angiography/endoscopy.
  • When endoscopy has not identified a source.

How to read it

  • A focus of luminal activity that appears and moves antegrade/retrograde indicates active bleeding.
  • Note the time of appearance and the bowel segment.
  • Distinguish true bleeding from fixed vascular/physiological activity.

Protocol

  • In vitro red-cell labelling (kit, ≥95% efficiency) is the method of choice; modified in vivo (~90%) is the alternative; 555–1,110 MBq.
  • Continuous dynamic imaging (frames no longer than 60 s; 10–20 s frames localise better) for at least 60 min if negative, as long as practical; delayed images are optional and localise less accurately; SPECT/CT can help localise.
  • Cine review to track movement of extravasated tracer.

Pitfalls

  • Localises a region, not the precise vessel.
  • Free pertechnetate (poor labelling) mimics bleeding (gastric/urinary activity) — image the neck: thyroid/salivary activity confirms free pertechnetate.
  • Rapid transit can mislocalise the origin.
Evidence & guidelines
  • Labelled-RBC scintigraphy is more sensitive than angiography for slow bleeding.
  • SNMMI/ACNM GI bleeding scintigraphy procedure standard 3.0 (Dam et al., J Nucl Med Technol 2025;53:14–23).
In depth
  • Labelled-RBC scintigraphy can detect bleeding as slow as 0.04 mL/min in animal models and about 0.1 mL/min clinically, well below the roughly 0.5 mL/min usually needed for catheter angiography.
  • ⁹⁹ᵐTc-labelled RBCs are the agent of choice: in a prospective tandem study they had 93% sensitivity, 95% specificity and 94% accuracy, whereas ⁹⁹ᵐTc-sulfur colloid detected only 5 bleeding sites (sensitivity about 12%).
  • RBC labelling efficiency is highest with the in vitro kit method (≥95%, method of choice), about 90% with the modified in vivo method, and lowest with the in vivo method, which is not recommended because of more free pertechnetate.
  • Imaging: 128×128 matrix, an optional angiographic phase at 1 frame per 1–3 s for 60 s, then continuous dynamic imaging with frames no longer than 60 s (10–20 s frames aid localisation) for at least 60 min, with delayed imaging possible up to 24 h.
  • A positive study needs all three criteria: activity appearing outside the expected blood-pool structures, a change in intensity on consecutive images, and movement in a pattern consistent with bowel (antegrade or retrograde).
  • Report the origin as the site of initial activity (not the most intense/largest); small-bowel bleeds move rapidly and centrally, large-bowel bleeds drape peripherally.
  • Pitfalls: free pertechnetate (gastric/thyroid/salivary), ectopic/transplant kidney, varices, accessory spleen, post-operative hyperaemia.
  • Meckel scan (⁹⁹ᵐTc-pertechnetate for ectopic gastric mucosa) has an overall sensitivity of about 85%, specificity 95% and accuracy 90%, largely from paediatric series, and is less sensitive in adults; the ectopic focus appears at the same time as gastric activity, which is most prominent after 10–15 min.
  • Meckel preparation: fasting 3–4 h is helpful but not mandatory; avoid recent barium studies, stop GI-irritating drugs or procedures (e.g. laxatives, enemas) for 2–3 days, and avoid perchlorate, which blocks mucosal uptake; H₂ blockers or proton pump inhibitors may improve sensitivity.

Sources: SNMMI/EANM GI bleeding scintigraphy 2.0 (PMID 25472517) · Thorne 1987 (PMID 3494826) · Nusbaum & Baum 1963 (PMID 14065652) · Bunker et al. AJR 1984 (PMID 6331732) · SNMMI/EANM GI bleeding scintigraphy 2.0 (PMID 25472517) and 3.0 (PMID 39909579) · SNMMI/EANM Meckel scintigraphy guideline 2.0 (PMID 24948825) · Sfakianakis 1981 (PMID 6265609)

Practise with a case
← BackGastrointestinal