Which nuclear medicine procedures are useful in localizing lower GI bleeding?
The difficulty of localizing acute lower GI bleeding is well recognized. Even acute and rapid bleeding can be intermittent and not detected on angiography. Alternatively, the culprit lesion can be obscured by luminal blood during endoscopy. Small bowel bleeding distal to areas accessible by upper endoscopy is notoriously difficult to localize.
Two nuclear procedures have been used to localize GI bleeding sources: short-term imaging with 99m Tc-sulfur colloid injection and extended imaging using 99m Tc-tagged RBC injection. Despite the theoretical advantage of 99m Tc-sulfur colloid in being able to detect smaller bleeds, this technique shares the limitation of angiography: a short intravascular residence time, which mandates the bleed to be active at the exact point of imaging. In addition, the normal biodistribution of sulfur colloid to the liver and spleen limits the evaluation of possible bleeds around the hepatic and splenic flexures. 99m Tc-RBC imaging has assumed dominance because the long intravascular residence time allows detection of intraluminal radioactive blood accumulation if extended imaging is necessary.
The first step is performing an in vitro tagging of RBCs with 99m Tc-pertechnetate, which provides the highest RBC tagging efficiency. In vitro tagging of radiolabeled RBCs involves obtaining a small blood sample (1 to 3 mL) from the patient and using 99m Tc-pertechnetate to label the RBCs in reaction vials. The radiolabeled RBCs are injected back into the patient and dynamic 1- or 2-second flow images are obtained in 60 seconds. In the case of a brisk bleed, the flow images will allow for better localization because delayed images will demonstrate significant radiotracer spread through the bowel. Immediately after dynamic flow images are obtained, sequential 1-minute images are acquired for 90 minutes. The use of dynamic imaging is important because sensitivity for localization is higher when the study is displayed in a cine-loop. If the patient has an intermittent bleed and the initial study is negative, images can be acquired up to 24 hours later if the patient actively bleeds again without reinjecting additional tagged RBCs. Unfortunately, delayed images will have a significant disadvantage in localizing the area of active bleeding because of normal peristaltic activity and the additional time from the beginning of the bleed to the time of imaging.