Oncology: F-18 Fluorodeoxyglucose Positron Emission Tomography

Physical Properties

In positron radioactive decay, a positron ( β ⁺ ) ejected from the atom travels a short distance before meeting a negative particle (electron) and undergoing annihilation. The resulting two 511-keV photons travel at 180 degrees from each other. These high-energy photons do not interact well with routine gamma cameras but are optimally detected by the specialized ring of detectors in a PET camera. Photons received within a short enough time interval at opposing detectors are registered as “coincidence photons,” or those originating from the same decay event. The result is a superior image to those achieved with gamma camera single-photon studies (e.g., with technetium-99m–labeled agents).

Many PET-emitting isotopes have very short half-lives (T _1/2 ), requiring a cyclotron to be in extremely close proximity. The 109.7-minute T _1/2 of F-18 means it can be shipped from local and even regional production facilities. On the other hand, the T _1/2 is not excessively long, and radiation exposure is lower than many agents with longer-lived radiolabels. Dosimetry information is outlined in Appendix 1 .

Kinetics and Distribution

In malignant cells, increased expression of membrane glucose transporters (e.g., glut-1) results in higher levels of intracellular glucose. Within these cells, the levels of hexokinase (hexokinase II) activity are also increased, phosphorylating glucose, which then moves through the glycolysis pathway. F-18 FDG is taken into the cell and phosphorylated by the same mechanisms as glucose, but F-18 FDG cannot be metabolized further ( Fig. 12.1 ). In addition, due to the lower levels of glucose-6-phosphatase in cancer cells, FDG remains effectively trapped because phosphorylated FDG cannot diffuse back across the membrane. The normal distribution of F-18 FDG is shown in Figs. 12.2 and 12.3 .

Patient Preparation

Because glucose competes with F-18 FDG for uptake, the patient’s glucose level should be checked before injection. The upper-limit cutoff varies among institutions, but a value under 200 mg/dL is generally considered acceptable.

Insulin, whether endogenously released in response to a meal or following administration to diabetics, stimulates the glucose transporters (e.g., glut-1), which are highly expressed in muscle cell membranes. This dramatically increases muscle uptake ( Fig. 12.4 ), thereby potentially decreasing uptake in tumor. To minimize the impact, patients fast overnight or for at least 4 to 6 hours before injection and avoid carbohydrates in the meal (or even the day) before injection. Patients with diabetes should not be given short-acting insulin within 2 hours of radiotracer injection, and long-acting insulin should be held overnight. The scan must be carefully scheduled because it may be difficult to coordinate with the diabetic patient’s serum glucose levels, which can fluctuate widely over the course of the day. When this is the case, optimal times often include early morning, before eating or taking insulin, and early afternoon, with the patient fasting after eating a light early breakfast and taking the morning short-acting insulin dose. For non–insulin-dependent diabetics taking the medication metformin, consideration should be given to withholding it for 1 to 2 days because metformin has been shown to dramatically increase uptake in the bowel.

Muscle activity is also minimized by limiting vigorous exercise for 1 to 2 days before the examination, and sedatives (e.g., alprazolam oral 0.5 mg) are routinely administered for patients with head and neck cancer who have undergone surgery in the past, helping to prevent frequently problematic increased background muscle activity.

Dose Administration and Uptake

Administered doses of F-18 FDG have been steadily decreased over the years. With the introduction of improved camera technology, such as time-of-flight (TOF) detection, typical doses are on the order of 7 to 8 mCi (259–296 MBq), up to half that used with the previous-generation scanners.

Patients should remain completely still and quiet for the uptake period, which is usually 50 to 65 minutes to achieve the optimum target-to-background ratio balanced with the physical decay T _1/2 . Some studies have suggested that additional delayed images at 90 to 120 minutes may improve sensitivity and specificity because tumors tend to continue accumulating F-18 FDG while activity continues to decrease from other tissues and benign processes. In the case of astrocytomas, which are especially difficult to image with F-18 FDG, an even greater interval (on the order of hours) may increase the accuracy. However, prolonged waiting periods are not generally practical in the clinic. Whatever delay is used, subsequent studies should be performed in a consistent manner to be certain that the changes seen are not artificially created.

PET/CT Scan Acquisition

Patients are usually imaged in the supine position after bladder voiding. Because CT artifact occurs when the arms are in the field of view, they are most often placed above the head when the pathology is in the chest, abdomen, and pelvis but left at the patient’s side when the tumor is in the head and neck.

The study is acquired in two phases. First, a transmission scan is performed using an external radiation source for attenuation correction. Originally, radiation sources (e.g., germanium-68 or cesium-137 rods) rotated around the patient, requiring several minutes. CT radiography, on the other hand, requires only seconds to cover the entire body. Based on the interactions of the x-ray photons with tissues, an attenuation-correction map is built and applied to the photons detected from the patient during the second phase of imaging, the emission PET scan. Attenuation correction allows PET data to be displayed with the proper intensity and is necessary for quantification of activity. The emission scan is acquired as a series of partially overlapping blocks of data, or bed positions, as the patient is moved through the camera. The scan time for each bed position can be modified depending on the patient’s body habitus but is typically a couple of minutes on modern TOF scanners, nearly half that of previous-generation scanners. Traditionally, a whole-body PET/CT refers to a scan extending from skull base to midthigh, although studies can include the extremities and brain, depending on the situation.

Although most PET/CT studies are performed without intravenous CT contrast, its use has been increasing because it helps identify normal structures and makes pathology more conspicuous. Water can be used to distend the stomach and duodenum. Dilute oral or water-equivalent negative oral contrast are also acceptable. Attenuation correction can help overcome any questions or artifacts if dense contrast builds up, causing artificially elevated counts.

Displays include a three-dimensional maximum-intensity projection (MIP) image and sagittal, coronal, and axial fused and unfused slices. Software can fuse PET images to CT scans performed at other times or to an MR image if desired. If the examination was performed on a dedicated PET/MR camera, it is still advisable to compare the study to a recent CT in order to visualize some abnormalities, such as small lung nodules.

Pharyngeal and Parapharyngeal

Mild activity is normally present in the salivary glands. Marked uptake is often seen in oropharyngeal lymphoid tissue, including the palatine and lingual tonsils. Asymmetry can occur normally or as the result of therapy and inflammation but may make the evaluation for tumor more difficult. Diffuse marked uptake is often seen in the vocal cords and oropharynx and is increased from speaking. In vocal cord paralysis, unilateral uptake may occur in the normal vocal cord, along with decreased activity in the contralateral abnormal vocal cord ( Fig. 12.6 ).

Myocardial Metabolism

Minimizing cardiac activity is desirable when evaluating cancer. The myocardium uses glucose as an optional fuel source. In a fasting state, fatty acid metabolism dominates over glycolysis, leading to decreased FDG uptake. However, fasting yields inconsistent results, and significant cardiac uptake is seen in up to 50% of fasting patients, often very heterogeneous in the left ventricle. Glucose loading, such as done for a cardiac viability study, increases glycolysis and, therefore, FDG uptake. Benign, fairly intense activity is occasionally seen in the intraatrial septal fat.

Urinary Excretion

Activity in excreted urine can create interpretation difficulties. Although the ureters usually appear as long, tubular structures, they can be seen as very focal areas of activity that may be confused with tumor or lymph node metastasis. Correlation with CT and MIP images can help confirm ureter activity or rule out a soft tissue lesion. In addition, activity in the filling bladder can obscure lesions in the pelvis, and urine contamination on the skin may be difficult to differentiate from an actual superficial lesion.

Gastrointestinal Tract

The esophagus normally shows no significant uptake. Nonspecific focal or more diffuse activity could be due to inflammation from reflux. In general, this is less than that seen with cancer or acute radiation. Significant activity in the stomach, especially when collapsed, can limit the usefulness of F-18 FDG in the evaluation of gastric adenocarcinoma or lymphoma. Highly variable activity in both the small and large bowel is especially problematic because it may obscure tumor in the bowel and mesentery. Focal transient activity is frequently seen in the small bowel, possibly the result of lymphoid tissue uptake. Marked uptake in the bowel can occur from metformin. Inflammation, such as from colitis, inflammatory bowel disease, appendicitis, and diverticulitis, can cause radiotracer accumulation ( Fig. 12.7 ). However, these diseases should cause corresponding changes on CT: fat stranding, inflammatory fluid collections, changes in the bowel wall, and air within forming abscesses.

Reproductive Organs

Cyclical changes can be seen normally in the uterus and ovaries in premenopausal women ( Fig. 12.8 ). Uniform, diffuse, mild to moderate endometrial uptake normally occurs, with maximal uptake occurring during menstruation (menstrual cycle days 0–4) and near ovulation (approximately day 14). Activity is lower during proliferative (days 7–13) and secretory (days 15–28) phases. Normal activity can occur in the ovaries related to ovulation and occasionally in association with follicle growth and the development of the corpus luteum cyst. Increased F-18 FDG uptake can be seen in benign uterine leiomyomas and endometrioma. In men, the testes vary widely but often show high-F-18 FDG levels normally.

Thyroid Disease

Thyroid activity is normally low or absent; however, multiple uptake patterns can occur ( Fig. 12.9 ). Diffusely increased uptake may be seen in thyroiditis, radiation thyroiditis, and Graves’ disease. The significance of low-level diffuse activity in patients without identifiable thyroid disease is uncertain; it may be normal or the result of subclinical thyroiditis. Focal uptake can be seen in benign adenomatous nodules. However, focal activity can be the result of malignancy in 30% to 50% of cases, and evaluation with ultrasound is warranted to determine whether a biopsy is needed.

Brown Adipose Tissue Activation

Brown fat (or BAT) plays a role in nonshivering heat generation and can be stimulated by the adrenergic system as well as the cold ( Fig. 12.10 ). It is particularly important in the young but is also occasionally seen in adults. When BAT stimulation is present, F-18 FDG uptake is seen in the fat of the supraclavicular region and neck and occasionally in the upper mediastinum and suprarenal regions. Uptake frequently occurs bilaterally in the region of costovertebral junctions. Sedatives (e.g., lorazepam or diazepam) and beta-adrenergic blockers (e.g., 20 mg oral propranolol) are sometimes used to decrease this uptake, although the effectiveness is variable. Rather, it is generally preferred to keep the patient warm for a period before injection until the examination is complete because this tends to be more effective. In some cases, a scan may have to be repeated because it may be difficult to rule out lymph node involvement such as from lymphoma ( Fig. 12.11 ).