Indications and Contraindications

A Tc 99m–radiolabeled erythrocyte scan should be performed to help localize the vascular origin of acute gastrointestinal bleeding involving the colon. The technique is less accurate in the assessment of upper gastrointestinal tract bleeding. The only major contraindication is a bleed so rapid that it is best spared scintigraphic interrogation in favor of immediate contrast angiography. The bleeding localization study may require too much radiolabeling preparation and imaging time to be performed on an acutely unstable patient.

Technique Description

No specific patient preparation is required for a Tc 99m–radiolabeled erythrocyte scan. This is an advantage because studies for gastrointestinal bleeding are usually requested emergently. Various labeling methods and kits are available to label the red blood cell, including an in vitro method, an in vivo method, and a hybrid method. The in vitro method using widely available kits typically provides very high labeling efficiency (e.g., Ultratag kit [Mallinckrodt, Inc., St. Louis]) provides a labeling efficiency of 98.5%).³ What the methods have in common is the use of stannous ion as a reducing agent and the Tc 99m label.

The patient is positioned supine with the camera located anteriorly. On initial intravenous injection of the radiolabeled erythrocytes, initial flow images are obtained at 1 s/frame for 60 frames. Subsequent dynamic images are obtained at 1 min/frame for 60 additional frames. A 128 × 128 image matrix is normally used for both acquisitions.⁴ Additional imaging may be repeated at intervals up to 24 hours, but all acquisitions should be in a dynamic series to enable detection of subtle bleeding sites and peristaltic transit through the bowel.

Pitfalls and Solutions

The most common technical pitfall of the Tc 99m–radiolabeled erythrocyte scan relates to an imperfect tagging of the blood cells, leaving free Tc 99m pertechnetate in the bloodstream. As described earlier, tagging efficiency is typically very high, especially with the in vitro technique. Common drugs such as heparin, penicillin, and iodinated contrast media interfere with the entry of the reducing agent (Sn⁺⁺) through the red blood cell membrane, however, causing an increased amount of free pertechnetate in the blood.⁵ Also, alternative in vivo labeling techniques may allow a certain amount of free pertechnetate to circulate.

Free pertechnetate appears in the stomach where it is physiologically secreted by the gastric parietal cells (see Fig. 85-2). This activity has an intraluminal configuration and moves antegrade over time. This movement can simulate an upper gastrointestinal bleed. In some instances, an upper gastrointestinal bleed may have been ruled out from a recent endoscopy, or the clinical presentation may not be consistent with one. A static image of the neck also reveals physiologic uptake of pertechnetate in the thyroid gland. After confirmation of the presence of free pertechnetate, the best solution is to allow for it in the interpretation. A brisk lower gastrointestinal bleed should be differentiated easily on dynamic imaging.

Another pitfall, related to the physiology of gastrointestinal bleeding itself, is intermittent bleeding. In Tc 99m sulfur colloid scanning (no longer commonly performed), after an initial 20-minute window, the study ceases to be diagnostic because of rapid radiopharmaceutical clearance by the liver. With Tc 99m–radiolabeled erythrocyte scans, however, delayed imaging up to 24 hours is possible. Delayed imaging can solve the problem of bleeding that has stopped temporarily during the initial imaging session.

Several factors must be accounted for to maximize the diagnostic utility of delayed scanning. First, the patient should be rescanned as soon as possible after the repeat bleed comes to clinical attention. The images obtained should be dynamic; this is the only way to localize the bleed. Localization is often impaired because the intraluminal blood, a strong peristaltic stimulant, has progressed distal to the actual bleeding site.

Interpretive pitfalls consist of confounding patterns of activity. The only solution for these interpretive pitfalls is for the interpreter to be aware of their appearance. Physiologic penile blood flow may be mistaken for rectal bleeding, or mesenteric vascular activity may overlie the expected location of the bowel. What these confounding patterns have in common is their lack of intraluminal configuration and the absence of antegrade or retrograde movement over time. Table 85-1 presents examples of these patterns.

TABLE 85-1 Confounding Patterns of Activity

Image Interpretation

Indications and Contraindications

The Tc 99m–pertechnetate scan is useful in detecting a Meckel diverticulum, which contains ectopic gastric mucosa whose acid production ulcerates the small bowel mucosa. These congenital abnormalities are almost always located in the right lower quadrant of the abdomen and manifest with symptomatic gastrointestinal bleeding (see Fig. 85-2). In 50% of cases, diverticula manifest before 2 years of age. The pertechnetate administered crosses the placenta,⁶ and women of reproductive age should be questioned regarding the possibility of pregnancy.

Technique Description

Scanning for the gastric mucosa in a Meckel diverticulum (also known as Meckel scan) should be obtained with the patient fasting for 4 hours before the scan. The camera is placed in the anterior position relative to the patient, who is positioned semisupine with left side down and right side up; this maneuver minimizes antegrade passage of physiologically secreted gastric pertechnetate activity from the stomach to the proximal small bowel. The field of view is kept with the xiphoid process superiorly and the pubic bone inferiorly, although more of the torso and extremities is inevitably included with very young pediatric patients.

After appropriate positioning, 5 to 10 mCi of Tc 99m pertechnetate is injected intravenously. The dose may be administered by weight for pediatric patients. A standard protocol includes flow images, obtained at 1 s/frame for 60 frames. Dynamic images are obtained at 30 s/frame for 60 additional frames. A 128 × 128 matrix is used.⁴

Pitfalls and Solutions

Technical pitfalls in Meckel scanning are rare because the radiotracer is simple to prepare. Because pertechnetate is secreted by the gastric mucosa in a Meckel diverticulum, most Meckel diverticula that do not contain gastric mucosa are not detected by the scan. Symptomatic small bowel bleeding is caused by only a few Meckel diverticula that do contain gastric mucosa, however. The physiology of radiotracer localization ensures that the scan detects only symptomatic diverticula—a limitation that is, from the surgeon’s standpoint, an advantage.

The most common interpretive pitfall relates to the physiologic excretion of pertechnetate through the renal system. Renal collecting system activity can be present in the expected location of a Meckel diverticulum. A lateral view reliably distinguishes the retroperitoneal structure.

Image Interpretation

Indications and Contraindications

Acetazolamide-augmented brain perfusion SPECT scans are obtained in patients with compromised carotid circulation who must be evaluated for external-to-internal carotid arterial bypass. If there is appropriate augmentation of cerebral blood flow from baseline after administration of the vasodilator, the cerebrovascular reserve is intact. If there is a failure of perfusion augmentation, this implies a critically stenotic vessel with compromised ability to increase regional blood flow to the brain parenchyma it supplies (Fig. 85-3).

FIGURE 85-3 Juxtaposition and software subtraction of Tc 99m HMPAO SPECT images obtained before (top row) and after (bottom row) administration of acetazolamide shows failure of augmentation in the right frontal lobe on the study obtained after acetazolamide administration (arrow, second row), indicating impaired cerebrovascular reserve.

The main contraindication to this study is a sulfa drug allergy. As with any drug allergy, a full investigation into the nature and severity of a past reaction should be obtained. Isolated reports of “steal phenomenon” leading to ischemia exist in the literature,⁷ but there are no data to suggest a significant risk of stroke after administration of this carbonic anhydrase inhibitor. Patients should be advised regarding side effects such as nausea, numbness around the mouth or fingers, lightheadedness, or facial flushing.

Technique Description

The patient should be placed in a dark, quiet room for 20 minutes before injection of radiotracer. The patient should not read or speak. Approximately 20 minutes after intravenous injection of 10 mCi of a Tc 99m–radiolabeled lipophilic amine compound—Tc99m HMPAO (Ceretec) or Tc 99m ECD (Neurolite)—a baseline SPECT scan should be acquired. Subsequently, 1 g of acetazolamide is infused intravenously over 5 minutes. Ten minutes after this, a higher dose (30 mCi) of the same Tc 99m radiotracer is administered. Another 20-minute delay precedes the second SPECT image acquisition.⁸

Pitfalls and Solutions

Patient motion is a common technical pitfall with brain SPECT imaging. Postprocessing software is often able to correct for a certain degree of patient motion. If patient stimulation is not minimized before the injection of radiotracer for the baseline and postacetazolamide studies, cortical activation can lead to spurious radiotracer localization. Brain SPECT with a lipophilic amine tracer detects brain perfusion, and brain ¹⁸FDG-PET detects metabolism; these are intimately related physiologic processes that can be artifactually altered if the patient is not kept in a dark room with minimal auditory stimuli.

Postprocessing errors can confound the interpretation as well. Many software programs compare the acquired data with a probabilistic brain atlas.⁹ Any results obtained through these means should be verified with direct visual inspection and used at best as an adjunct corroborating the interpreter’s own impression.

Interpretive pitfalls are rare with brain SPECT imaging. The vascular changes visualized are often significant ones, in contrast to the more subtle alterations in perfusion seen with interictal and ictal SPECT scans for epilepsy evaluation.

Image Interpretation

Reporting

Areas of absent or severely limited cerebrovascular reserve appear as areas of decreased perfusion compared with baseline after vasodilator administration (Fig. 85-4). The report should list these areas, ideally taking advantage of digital image fusion or coregistration with correlative anatomic studies, such as MRI or CT.

FIGURE 85-4 Tc 99m HMPAO SPECT shows severely impaired perfusion to the left cerebral hemisphere, especially the left parieto-occipital lobe and frontal lobe (arrows). The absence of change between studies before and after acetazolamide administration indicates an absence of cerebrovascular reserve.

KEY POINTS

Tc 99m–radiolabeled erythrocyte scans guide selective angiographic intervention in the setting of acute lower gastrointestinal tract bleeding.

Tc 99m pertechnetate scanning detects the ectopic gastric mucosa that is present in symptomatic Meckel diverticula.

Patients undergoing vasodilator-augmented brain SPECT imaging for assessment of cerebrovascular reserve benefit from advanced software analysis for comparison of baseline versus augmented scans and comparison with an atlas of established normals.

As with all functional imaging studies, an understanding of the physiology to be interrogated, image data acquisition, and postprocessing techniques optimizes the final interpretation.