The Skeletal System

Preparation

Tc-99m MDP can be prepared from a simple kit. Tc-99m, in the form of sodium pertechnetate (Na ^99m TcO ₄ ) obtained from the technetium-molybdenum generator, is injected into a vial containing MDP, stabilizers, and stannous ion. Stannous tin (Sn II) acts as a reducing agent, allowing the Tc-99m pertechnetate to form a chelate bond with the MDP carrier molecule.

Incomplete labeling may occur if air is introduced into the vial because oxygen causes hydrolysis of the stannous ion (from Sn II to Sn IV). Insufficient stannous ion results in free technetium pertechnetate (“free tech”), causing image degradation with increased background soft tissue activity and uptake in the thyroid, stomach, and salivary glands. Tc-99m MDP should be used within 2 to 3 hours of preparation, or radiopharmaceutical breakdown may also yield free technetium pertechnetate. Excess alumina from the technetium generator eluate may lead to colloid formation, which can be seen as uptake in the reticuloendothelial system of the liver.

Uptake and Pharmacokinetics

The injected Tc-99m MDP rapidly distributes into the extracellular fluid and is quickly taken up into bone. Although accumulation relates to the amount of blood flow to a region, uptake is primarily the result of osteogenic activity, being much higher in areas of active bone formation and repair than in mature bone ( Fig. 6.3 ). Tc-99m MDP binding occurs by chemisorption in the hydroxyapatite mineral component of the osseous matrix. Accumulation in areas of amorphous calcium phosphate may account for the Tc-99m MDP uptake sometimes seen in sites outside the bone, such as dystrophic soft tissue ossification. Decreased activity is seen in areas of reduced or absent blood flow or infarction. Diminished uptake or cold areas are also often seen in lytic metastases.

Approximately 50% of the dose is localized to the bone, with the remainder excreted by the kidneys. Although peak bone uptake occurs approximately 1 hour after injection, the highest target-to-background ratios are seen after 6 to 12 hours. Images are typically taken at approximately 3 hours to balance the need for background clearance with the relatively short 6-hour half-life of Tc-99m and patient convenience. Also, the radiotracer half-life limits imaging to a maximum of 24 hours after injection. Radiopharmaceutical dosimetry is discussed in Appendix 1 .

Imaging Protocol of Tc-99m MDP

The patient should be well hydrated and, after injection, should be instructed to drink several cups of fluid to improve background clearance. Frequent bladder voiding reduces the radiation dose. Care must be used because urinary contamination can cause confusion or mask potential lesion sites.

There are three main phases that can be imaged with a bone scan. The first phase involves rapid dynamic image acquisition immediately after injection to assess blood flow to an area of concern. This can be followed by the second, soft tissue, phase, which lasts a few minutes. Delayed images that visualize the bones make up the third phase. Occasionally, further delay, a so-called fourth phase, is still needed at 18 to 24 hours to clear soft tissue activity and maximize target-to-background ratios.

Before injection, a decision must be made as to whether or not blood flow and soft tissues need to be evaluated as part of a three-phase examination or if routine delayed images alone will suffice. Box 6.1 lists uses for both acquisition protocols. Metastatic disease and back pain assessment can be done by delayed imaging alone. The presence of increased arterial blood flow and abnormal soft tissue activity can aid in the diagnosis of several additional acute problems: osteomyelitis, painful joint prosthesis, fracture, avascular necrosis (or osteonecrosis), bone graft status, and complex regional pain syndrome (previously called reflex sympathetic dystrophy). A sample protocol for three-phase and routine delayed scanning is listed in Box 6.2 , and an example of the parameters used for SPECT imaging is summarized in Box 6.3 .

If dynamic three-phase scanning is to be performed, a bolus of Tc-99m MDP is injected intravenously with the area in question under the camera ( Fig. 6.4 ). The injection site should be chosen to avoid any suspected pathological condition. For example, if comparison with the opposite hand may be needed at any time, injection in a site such as the foot should be considered. The first phase consists of serial 1- to 4-second dynamic blood flow images acquired for 60 seconds. Then blood pool or soft tissue second-phase images are obtained of the main region and any secondary areas of interest, such as in patients with arthritis or multiple stress injuries. The timing of delayed images may vary, from 2 hours in younger patients up to 3 to 4 hours in the elderly, obese, and in those with poor renal function.

Using a low-energy, high-resolution collimator, delayed planar images can be obtained by whole-body scan or spot views. The whole-body scan allows rapid, seamless coverage as the camera moves over the patient at a predetermined rate. On the other hand, spot views, with the camera fixed over each area to be imaged, provide greater resolution and detail. In most centers, a whole-body scan is performed, with high-count spot views reserved for symptomatic areas or additional views of suspicious-appearing regions ( Fig. 6.5 ).

Other modifications can be performed. Magnified views with a pinhole collimator can better visualize joints in children, osteonecrosis of the hips, and trauma to the carpal bones. In SPECT, a volume of data is acquired from camera detectors orbiting the area of interest, and the reconstructed 3D images can be formatted in transaxial, sagittal, and coronal planes. SPECT images provide better resolution, more precise localization, and improved contrast of cold and hot lesions ( Fig. 6.6 ). Not only can this increase sensitivity, but more exact localization can increase specificity by avoiding confusion with benign uptake ( Fig. 6.7 ).

Correlating findings with CT is often key to making the correct diagnosis. When no radiographic abnormality is seen to explain activity, suspicion is increased that uptake is the result of early metastasis. Although commercially available software can fuse the SPECT with a CT obtained at another time, studies acquired on a hybrid SPECT/CT scanner usually result in superior results because of the more uniform slice thickness and positioning between the two examinations and because of the elimination of the chance that the pathology will change in the interval between studies.

F-18 NaF Imaging Protocol

After intravenous injection, F-18 NaF localizes to areas of new bone formation through chemisorption, in a similar fashion to that of Tc-99m MDP. However, sodium fluoride is not highly protein bound in the blood, resulting in rapid renal clearance.This allows imaging to begin as early as 30 to 45 minutes (although a longer delay, 60-90 minutes, will result in superior images). Given the high target-to-background ratios and excellent resolution, PET images can be performed without CT attenuation correction. However, intensity is more uniform with attenuation correction, preventing bones in areas subject to less attenuation from falsely appearing more intense than those where photons are more likely absorbed or scattered. CT is also very useful to localize lesions to areas of bone pathology and differentiate malignancy from benign processes.

The low-dose whole-body CT requires only seconds. This is followed by the emission PET scan data, collected as the patient moves through the scanner in a series of bed positions, each covering several centimeters at 1 to 2 minutes per bed position. A few sites have advocated the use of a “cocktail” combining F-18 FDG and F-18 NaF in order to assess the bones and soft tissues in one session.

Patients should be well hydrated, void their bladders frequently, and continue to drink extra fluids for a few hours after injection to minimize radiation to the bladder. Although the higher energy of PET photons can lead to increased radiation doses to the patient, its relatively short half-life of 109.8 minutes helps limit exposure, which is outlined in Appendix 1 .

Normal and Altered Distribution

The appearance of the bones varies dramatically with age. Most notably, the growing skeleton will concentrate radiotracer at all active growth plates ( Fig. 6.9 ). These areas are also often the critical sites in trauma, primary bone tumors, and osteomyelitis. Therefore, it is essential that children are immobilized and positioned symmetrically in terms of rotation and distance from the camera face. By adulthood, growth-plate uptake diminishes and disappears. Normal activity can persist in some areas, such as residual ossification centers in the sternum and the sternomanubrial joint.

Some bones, such the sacroiliac joints, the iliac wings, or a lordotic spine, normally appear more intense because of their greater density or closer proximity to the camera. Increased uptake bilaterally in the frontal bones of the calvarium may occur from benign hyperostosis frontalis interna. Increased activity, sometimes asymmetrical, is occasionally seen where the sphenoid ridge meets the calvarium along the lateral orbits. The costochondral junction is another common site of benign uptake and is an unlikely location for metastasis unless uptake extends along the rib.

Osteoarthritic changes are routinely seen and usually identifiable by a classic distribution. Arthritis is frequently bilateral and often involves both sides of the joint. The areas typically involved are the spine, knee (particularly the medial compartment), feet, shoulder, wrist (especially at the base of the first metacarpal), ( Fig. 6.10 ). Uptake in the patella may result from chondromalacia and degenerative change. Mild asymmetry has been noted in the shoulders, apparently affected by handedness and use. Of note with F-18 FDG PET, arthritis rarely shows the high levels of uptake seen on bone scan. F-18 sodium fluoride, on the other hand, is often very abnormal in sites of arthritis and other benign lesions ( Fig. 6.11 ).

Assessment of spinal uptake frequently requires radiographic correlation with CT (or MR) in addition to SPECT. Some of the changes that can occur from degenerative arthritis include facet hypertrophy, disk space narrowing, osteophyte formation, and Schmorl’s nodes. Osteoporosis may result in classic insufficiency vertebral compression fractures ( Fig. 6.12 ). Abnormal uptake in the vertebra may be seen before radiographic changes occur and may not resolve, particularly in the elderly ( Fig. 6.13 ). Positive uptake has sometimes been used to help determine which patients might obtain symptomatic relief from the injection of bone cement (vertebroplasty). The H-shaped insufficiency fracture occurring in the sacrum ( Fig. 6.14 ) is frequently seen only on scintigraphic studies and not detectable on CT or MR.

The effects of trauma are often identifiable on bone scan. In the ribs, vertically aligned focal uptake in multiple, often successive ribs is a classic finding ( Fig. 6.15 ). Metastatic lesions, on the other hand, tend to track along the bone, as shown in Fig. 6.5 . When fractures are present, a poorly defined lytic lesion or aggressive periosteal change favors a pathological fracture, whereas regular callous formation is seen in a healing benign posttraumatic fracture. In some cases, the cause of a fracture may be difficult to determine without follow-up.

Soft Tissues

The appearance of the soft tissues on bone scan is of critical importance. Normally, the kidneys and bladder show excreted activity from radiotracer in the urine. Abnormal increased or decreased activity must be explained ( Box 6.4 ). Soft tissues (e.g., breast or abdominal fat) and implants attenuate the intensity of the underlying bones. Abnormal increased uptake outside of bone may be subtle or difficult to differentiate from true bone lesions.

In some instances, abnormal soft tissue uptake may be present from hemorrhage or necrosis, likely as a result of a combination of ossification and agent binding to macromolecules. The effects of recent surgery may be evident, and tumors (primary and metastatic) may be seen ( Figs. 6.16 and 6.17 ). In the right lower chest and upper abdomen, abnormal soft tissue activity may be a result of breast tumor, malignant pleural effusion, or metastatic adenocarcinoma of the breast or colon to the liver. Correlation with CT is helpful to uncover the disease that may be causing Tc-99m MDP accumulation ( Fig. 6.18 ). In addition to the effects of disease, abnormal activity in the liver may be the result of improper radiopharmaceutical preparation resulting in a colloid that is then taken up by the hepatic reticuloendothelial system. Radiopharmaceutical quality control tests can identify these cases.

When a three-phase scan technique is used, markedly increased perfusion indicates that the disease or trauma affecting the area is acute because chronic processes cause little or no asymmetry in blood flow, as shown in Fig. 6.4 . Increased blood flow and blood-pooling activity can result from cellulitis, abscess, or synovitis in the soft tissues ( Fig. 6.19 ). However, the changes can also occur as a result of disease in the underlying bone or joint, such as fracture or osteomyelitis.

Comparing Different Imaging Methods

Comparing the sensitivity of scintigraphic imaging with traditional modalities, it should be noted that bone scan requires as little as 5% bone turnover for a lesion to be detected, compared with the 50% needed on x-ray. Therefore lesions are often visualized on bone scan but not on x-ray. Bone scan is also usually more sensitive than CT.

MR is more sensitive than planar bone scan and potentially more specific. Its high-resolution images are often able to directly visualize disease-related signal changes in the marrow, and it can optimally evaluate soft tissue structures. However, when SPECT/CT is performed, the accuracy of MR and bone scan for skeletal disease is comparable. It is also difficult for MR to replace bone scan in several situations: obese patients and those who have claustrophobia, difficulty lying on the camera table for lengthy examinations, implanted metallic devices incompatible with the powerful magnetic gradients (e.g., pacemakers), intravenous contrast sensitivity, or renal insufficiency. The whole-body scanning desirable for metastatic disease evaluation is not widely available or practical at this time with MR.

Several small studies have shown that SPECT is more sensitive than planar bone-scan images and better able to localize lesions. The studies have also shown that the use of SPECT/CT improves specificity, leaving fewer indeterminate lesions in the end. As previously noted, images performed with PET are superior to studies from a gamma camera. Compared with bone scan, F-18 NaF has demonstrated higher lesion sensitivity overall, and F-18 FDG PET better detects aggressive, lytic lesions. The resolution of single-photon agents is lower than with PET/CT, but the use of SPECT/CT narrows or eliminates the gap in accuracy between the methods. In some cases, additional tests or short-interval follow-up will be needed to establish a diagnosis, whether the examination is performed with PET/CT or one of the bone-scan techniques.

Scintigraphic Patterns in Metastatic Disease

The scintigraphic patterns encountered in skeletal metastatic disease are summarized in Box 6.5 . Multiple focal lesions distributed randomly in the skeleton provide a high degree of clinical certainty in the diagnosis of metastases. However, other causes can also show multiple areas of uptake ( Box 6.6 ). Often, different features and patterns can help identify these causes. For example, Paget disease may be differentiated from metastasis by a coarse expansion of the bone.

The chance that a solitary lesion is a result of malignancy varies by location ( Table 6.1 ), and some features of primary bone lesions are outlined in Table 6.2 . Very focal uptake in one rib is likely from trauma. Even in a patient with known cancer, it has only a 10% to 20% chance of being malignant, whereas uptake in the central skeleton has a much higher likelihood of being from metastasis. Primary bone tumors, such as osteosarcoma, must be suspected, especially in younger patients, when long-bone involvement is seen. Common benign causes for a solitary lesion include arthritis and trauma. Some benign bone lesions, such as enchondroma, osteoma, fibrous dysplasia, osteomyelitis, and monostotic Paget disease, can also cause solitary abnormalities. Rarely, a benign bone island or a spinal hemangioma will accumulate some Tc-99m MDP. Stability over time can prove a lesion benign, and hemangiomas show a characteristic pattern on CT imaging, with striations or prominent trabecula.

• Blastic Metastases: Sites that are predominantly osteoblastic are more easily seen with Tc-99m MDP and F-18 NaF PET ( Fig. 6.20 ). With F-18 FDG PET, blastic lesions show variable activity, but uptake is often very low or at background ( Figs. 6.21 and 6.22 ). In response to treatment, lesions will show decreasing uptake, often occurring along with increasing sclerosis on CT imaging.

  
  
  

• Cold Lesions: Lesions that are aggressive, purely lytic, or completely replaced by tumor may show decreased uptake or appear “cold.” A list of possible causes for cold defects is provided in Box 6.7 . These photon-deficient areas may be difficult to spot because of overlying or adjacent activity, although SPECT and SPECT/CT increase sensitivity. In some cases, Tc-99m MDP bone scan and F-18 FDG PET/CT may be complementary tests, with blastic lesions often better seen with bone scan while F-18 FDG is superior for lytic areas. ( Fig. 6.23 ). With high sensitivity for lytic and blastic lesions, F-18 NaF PET/CT may also be useful in difficult cases. However, false negatives can occur, particularly in small lytic lesions in the spine. Other causes for decreased uptake such as metal attenuation artifact, radiation ports, compromised blood flow early in a pediatric septic joint, or very early infarct or avascular necrosis should be considered in the differential.

  Box 6.7

  Differential Diagnoses for Cold Lesions on Bone Scan

  • Metal attenuation artifact

  • Radiation changes

  • Barium in bowel

  • Early osteonecrosis (or avascular necrosis)

  • Early infarct

  • Multiple myeloma

  • Osseous metastases

    • Renal cell carcinoma

    • Thyroid cancer

    • Anaplastic tumors

    • Neuroblastoma

    • Breast and lung cancer (often mixed lytic and blastic)

  • Tumor involvement in marrow

    • Lymphoma

    • Leukemia

  • Benign tumor, cysts

  

• Superscan: A potentially problematic scintigraphic pattern is the “superscan” or “beautiful bone scan.” Fairly homogeneous increased activity is seen in the bones, with absent or only faint visualization of the kidneys and bladder (e.g., Fig. 6.3 , B ). The differential diagnosis of the pattern is provided in Box 6.8 , but most commonly, the cause is either diffuse prostate cancer metastases or metabolic change from severe renal failure causing hyperparathyroidism ( Fig. 6.24 ). Confusion from a superscan is a less common interpretive problem than in the past because of improved technology and image quality, and it should be pos sible to discriminate diffuse metastasis. Reviewing the available radiographs on each patient will help prevent mistakes.

  Box 6.8

  Causes for Superscan Pattern

  Common

  • Metastatic disease (especially prostate cancer)

  • Renal osteodystrophy

  • Prolonged delay prior to imaging

  Less Common

  • Severe hyperparathyroidism (rare in primary hyperparathyroidism)

  • Osteomalacia

  • Paget disease

  

• Flare Phenomenon: When patients have a good response to chemotherapy, they sometimes experience increased pain, and the bone scan may paradoxically worsen, with a flare of increased activity ( Fig. 6.25 ). If these lesions are followed over 2 to 6 months, CT imaging shows increased sclerosis from an osteoblastic response as the bone begins to heal. Activity should regress by 4 to 6 months after the flare. This phenomenon reinforces the fact that tracer uptake is not in the tumor but rather in the surrounding bone.

  

Prostate Carcinoma

Skeletal scintigraphy is generally best for the detection of prostate cancer metastases, and until the introduction of the prostate-specific antigen (PSA) blood test, bone scan was considered the most sensitive technique for detecting bone involvement. Serum alkaline phosphatase measurement detects only half the cases detected by scintigraphy. Radiographs may be normal 30% of the time.

The likelihood of an abnormal scintigram correlates with clinical stage, Gleason score, and PSA level. In early stage I disease, scintigrams demonstrate metastases less than 5% of the time. The incidence increases to 10% in stage II disease and 20% in stage III. In patients with PSA levels less than 10 ng/mL, bone metastases are rarely found (<1% of the time). Even if the risk is lower, scintigraphy is still indicated to evaluate symptomatic patients and suspicious areas seen radiographically. With increasing PSA levels, the chance of detecting metastatic disease increases. In cases where recurrent prostate cancer is a concern, bone scan may be used to supplement the detection of soft tissue involvement performed with F-18 fluciclovine (Axumin). Although F-18 NaF PET may be useful, F-18 FDG is most often not.