Bone metastases of solid tumors and secondary involvement of bone by hematologic malignancies are among the most frequent causes of tumoral involvement of bone.

Bone scintigraphy is a screening technique with a relatively poor specificity, because it studies the bone-forming reaction to the presence of the tumor but does not demonstrate the presence of tumor tissue itself. In addition, its sensitivity has been overestimated: bone scintigraphy is a relatively low-resolution technique and may miss small lesions or lesions that are purely lytic. An important point to improve the performance (particularly the specificity) of this technique is to ensure that reading is done by very experienced nuclear medicine specialists. Tomographic scanning by using SPECT and the combination with CT (SPECT-CT) may improve performance. Bone scintigraphy conversely has the advantage of being a whole-body screening method, with relatively low cost and low irradiation of the patient. It remains an important screening method, but needs control with a confirmatory examination
when a suggestive abnormality is described. Most often, these confirmatory methods will be plain radiographs for the appendicular skeleton and x-ray CT for the axial skeleton, particularly the spine.

[¹⁸F]fluoride positron emission tomography (PET) has been advocated as a more sensitive method. Spatial resolution and contrast-to-noise ratio are far better than with ⁹⁹Tc bone scintigraphy, but the method has a higher cost and requires use of a PET system that could be better used for true metabolic studies.

FDG-PET has the major advantage of imaging the tumor tissue, and not the sclerotic reaction of bone. It is based on the glucose avidity of the tumor due to increased metabolism. It is a whole-body method that detects not only bone lesions but also soft-tissue metastases. However, not all tumors in the bone are hypermetabolic. Prostate cancer is known to be poorly glucose avid, as are kidney cancer and some breast cancers before the tumor becomes more aggressive. PET-CT adds the localization power of x-ray CT and the time benefit of doing the attenuation correction by using CT data. PET is the best tool to detect bone lesions of lymphoma and lung cancer at staging or restaging. It is, however, under active evaluation for other tumors, notably advanced breast cancer, and may also be interesting in detecting early response to therapy.

Plain radiographs remained for many years the basic method of confirmation and follow-up method for bone metastases. As a screening method, it was used under the form of a “skeletal survey” including the bones that were statistically most often involved: the axial skeleton and proximal appendicular bones. X-ray tomography was used in case of unclear findings. These methods have no place today.

The sensitivity of plain x-ray is poor (1). The detection is late when the lesion either disrupts the bone cortex, destroys more than 30% of trabecular bone (in a vertebral body), or is accompanied by a significant sclerotic bone reaction.

During the follow-up, plain radiographs do not report with detail the evolution of bone lesions: they may mimic disease worsening in case of sclerotic response to therapy or dissimulate a new lytic evolution in a densified bone.

Plain radiographs disappeared from most of the diagnostic algorithms for secondary bone lesions. They retain, however, some utility as a quick method to confirm complications such as a vertebral collapse in a patient with known bone metastases.

Computed tomography (CT) is a high-resolution, x-ray tomographic technique that permits a detailed analysis of the trabecular structure of bone and evaluation of the cortical integrity. The densitometric sensitivity is by far superior to that of the plain radiograph, and replacement of fat by tumor in the bone marrow may be directly seen, as well as the extension of the tumor in the soft tissues. New multidetector CT devices made it possible to investigate larger body parts in a short time, to freeze physiologic motion, and to reduce irradiation.

Magnetic resonance imaging (MRI) relies on magnetic properties of the hydrogen nucleus (proton) to produce images whose contrast is governed by multiple parameters (the most important being the relaxation times, T₁ and T₂). Images may be influenced by the different parameters, depending on the choice of the user. In most instances, several sequences with different weighting will be obtained in the same examination (see
Fig. 4.3). This leads to a tremendous increase in contrast between tissues in comparison with methods such as CT. The drawback is that the method is more difficult to achieve and interpret, is more operator dependant, and may lack specificity. Its sensitivity is high, and it is able to detect lesions at the stage of early invasion of the bone marrow, even before any destructive effect on the bone matrix.

Progress in MRI (multiple array coils, speed, and spatial resolution) allows consideration of covering larger body areas, or even the whole skeleton, with sequences that are sensitive to metastatic disease of bone. Diffusion contrast imaging (3) and dynamic contrast-enhanced MRI (DCE-MRI) (4) hold a promise for better specificity, particularly for differential diagnosis between tumoral and nontumoral vertebral compression fractures.