Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Solitary bone lesions and soft tissue masses are commonly encountered in general orthopedic practice. Emphasis in this chapter will be on the radiographic appearance of solitary bone lesions. However, advanced imaging patterns will also be discussed, especially in the setting in which radiographic imaging contributes little to the evaluation of such lesions. With lytic and blastic bone lesions, radiographs often provide a discrete differential diagnosis. In those instances, advanced medical imaging is used primarily for staging and treatment planning. We will be focusing our discussion chiefly on the initial evaluation of these lesions and will avoid discussing follow-up imaging after treatment.
This section discusses the characteristic imaging findings of lytic and sclerotic bone lesions. We will also discuss the general approach to evaluating bone lesions. In the imaging section we will describe those tumors that can be confidently diagnosed by radiography, computed tomography (CT), and/or magnetic resonance imaging (MRI).
One of the primary functions of radiography is to detect lytic lesions in patients with nonspecific symptoms. It is important to be aware of normal variants that can mimic lytic bone lesions in order to avoid unnecessary biopsies. One of the most common variants that simulates a lytic lesion is the pseudocyst. A pseudocyst is an area of focal trabecular rarefaction within the bone at a location of low stress. When the area of rarefaction is visually compared to the surrounding bone that contains more prominent trabeculae, an apparent lytic lesion (or so-called pseudocyst) stands out. Fortunately, these areas of low stress that cause trabecular rarefaction are well described in the literature, and radiologists are familiar with these normal variants. Pseudocysts most commonly occur in the greater tuberosity of the humerus, the anterior aspect of the calcaneus, and the radial tuberosity ( Fig. 2-1 ). If there is any confusion as to whether there is a superimposed lytic mass over an area of a pseudocyst, an MRI can be performed. When only a pseudocyst is present, the MRI will show normal bone marrow signal in the area of concern.
After a true lytic lesion is identified, several imaging patterns or characteristics should be evaluated. These patterns are evaluated to determine if the lesion represents an aggressive or nonaggressive process and to narrow the differential diagnosis of the lesion. These patterns include the following: a periosteal reaction (evaluates aggressiveness), a zone of transition (evaluates aggressiveness), the presence of matrix mineralization (aids in differential diagnosis), and the location within the bone (aids in differential diagnosis). Finally, and probably most importantly, the patient's age should be considered along with these findings. The patient's age can be one of the most effective clinical tools in narrowing a large differential to only a few possibilities when considering nonspecific lytic lesions.
If a solid, uninterrupted periosteal reaction is present, then there is no destruction or transgression of the periosteum and the lesion can be regarded as nonaggressive ( Fig. 2-2 ). However, if a lesion does invade through the periosteum, this interrupted periosteal reaction suggests that the lesion is aggressive. Examples of an interrupted periosteal reaction would include a lamellated (known as onion-skinning) or perpendicular (sunburst) appearance, as well as Codman (tri)angle ( Fig. 2-3 ). An aggressive lytic lesion with an interrupted periosteal reaction represents typically an aggressive neoplasm or an aggressive pyogenic infection.
In the absence of a periosteal reaction, other imaging patterns must be used to determine if a lytic lesion is aggressive. If no periostitis is seen, then the lesion is either nonaggressive or it may be so aggressive that the bone has had little time to form a periosteal reaction ( Fig. 2-4 ). The most common nonaggressive lytic lesions that do not incite periostitis are fibrous dysplasia, enchondroma, nonossifying fibroma, and simple bone cyst. If any of these four lesions show periostitis, then the cause is usually a subtle pathologic fracture.
In medical imaging, the term zone of transition refers to the spatial relationship between diseased bone and normal bone at the border of a lytic lesion on radiographs or CT—but not MRI. A narrow zone of transition implies that there is normal-appearing bone immediately adjacent to the lytic lesion and that the precise border of the lesion can be readily seen on the radiograph. Such an appearance is said to be geographic ( Fig. 2-5 ). The differential diagnosis for common geographic lytic lesions is fairly small and includes chondromyxoid fibroma, nonossifying fibroma, and enchondroma. A narrow zone of transition indicates that the lesion is nonaggressive. A wide zone of transition implies that the border of the lesion is not as well defined and that there is a fair amount of bone destruction by the lesion along with tumor infiltration of the adjacent marrow. This type of bone destruction is described as moth-eaten ( Fig. 2-6 ) or permeative ( Fig. 2-7 ). Moth-eaten and permeative lesions are aggressive, with moth-eaten lesions tending to be less aggressive than permeative lesions. Differentiating the moth-eaten from the permeative pattern is not always easy, and the importance of differentiating these two patterns may be only “academic.”
There are other nuances to classifying the zone of transition using the method described by Lodwick. Using this method, the zone of transition is graded based on the aggressiveness of the lesion, with grade I lesions being primarily geographic, grade II lesions being moth-eaten, and grade III lesions being permeative. Grade I lesions are further subclassified into grades IA, IB, and IC. The subclassification of grade I lesions is beyond the scope of this chapter with the exception of grade IA lesions, which have less than a 2% likelihood of being malignant. Grade IA lesions are geographic, have a thick sclerotic rim, have no cortical penetration, and have no significant expansion (<1 cm of expansion in any plane) ( Fig. 2-8 ). If a grade IA lesion is asymptomatic and not characteristic for a definite tumor, it can be followed with serial radiographs.
Finally, the zone of transition in long bones and most flat bones can be readily evaluated on conventional radiographs. However, the zone of transition can sometimes be grossly misinterpreted on radiographs of anatomically complex osseous structures such as the scapula, ilium, ischium, or pubic bone. In such cases, the zone of transition is best evaluated by CT.
A lytic lesion can be considered clinically nonaggressive if it is painless, has a grade IA zone of transition, and has an absence of an interrupted periosteal reaction. All other lesions may be aggressive or exhibit aggressive clinical behavior. One should remember that a lesion that does not appear radiologically aggressive may still represent a malignant neoplasm and that an aggressive-appearing lesion may still be benign (e.g., osteomyelitis).
When mineralization within a lytic lesion is detected on radiographs, it indicates the presence of one of two types of tumor matrix. If the mineralization is arclike, ringlike, punctate, or popcorn-like, the lesion contains chondroid matrix ( Fig. 2-9 ). If the mineralization is wispy, cloudlike, or opaque, the lesion contains osteoid matrix ( Fig. 2-10 ). If the mineralization is too faint to characterize on radiographs, it can be better evaluated by CT.
Examples of tumors that may contain mineralization indicating a chondroid matrix include enchondroma, chondrosarcoma, and chondroblastoma. Some examples of tumors that may contain mineralization indicating an osteoid matrix include osteosarcoma and osteoblastoma.
Finally, a single dystrophic calcification located in the center of a lytic bone lesion is nearly pathognomonic for an intraosseous lipoma. Intraosseous lipomas are most commonly located in the anterior aspect of the calcaneus and in the long bones of the lower extremity ( Fig. 2-11 ).
The location of a bone lesion is important in narrowing the differential diagnosis. The most common nonaggressive-appearing lytic lesions occurring at the end of a long bone are subchondral cysts (if there is evidence of osteoarthritis), chondroblastoma (in patients younger than 30 years), giant cell tumor (if the physis is closed), metastasis (in patients older than 40 years or who have a known primary malignancy), multiple myeloma (in patients older than 40 years), and infection (if the physis is open).
The presence of multiple lytic lesions in different bones should raise the possibility of metastasis, multiple myeloma, infection, brown tumors (hyperparathyroidism), enchondromatosis (Ollier disease), or polyostotic fibrous dysplasia. The differential diagnosis can be further narrowed based on the age of the patient. For example, metastasis and multiple myeloma are the most common polyostotic lytic processes in patients older than 40 years; however, they are rare in adolescents and young adults.
The vast majority of geographic lytic lesions in the ribs are either fibrous dysplasia, metastasis, multiple myeloma, aneurysmal bone cyst, or Langerhans cell histiocytosis (eosinophilic granuloma). Again, this differential diagnosis can be further narrowed based on the age of the patient.
Lesions that are nearly always eccentric are nonossifying fibroma and giant cell tumor. This holds true to the extent that if a lesion is located centrally within a long bone, then nonossifying fibroma and giant cell tumor should not be highly considered. Conversely, a simple bone cyst is nearly always centrally located within a long bone (typically the proximal humerus and proximal femur); therefore, if a lytic lesion is eccentrically located within a long bone, a simple bone cyst should not be seriously considered.
Some tumors that present as lytic lesions have a predilection for specific bones. For example, approximately 85% of adamantinomas occur in the anterior cortex of the tibial diaphysis, whereas approximately 85% of chordomas occur in the sacrum or skull base.
MRI is primarily used for staging and treatment planning of malignant bone lesions. Depending on the pulse sequence used in a particular instrument, images are reported typically as being either bright or dark compared with isodense tissue in the same field. Typically, if a lesion is bright on a T1-weighted image, it represents protein-rich fluid such as blood or fat and occasionally may be highlighting calcifications. Alternatively, if the lesion is dark on a T1-weighted image but bright on a T2-weighted image, it signifies increased water content such as is present with edema, hemorrhage, or inflammation. If a lesion is dark on a T2-weighted image, it represents fibrous tissue, hemosiderin, or melanin. Occasionally MRI can contribute to the differential diagnosis. An example of this would be fluid–fluid levels within a lesion, which should raise the possibility of an aneurysmal bone cyst or, if there are aggressive features, a telangiectatic osteosarcoma ( Fig. 2-12 ). Additional information such as the presence or absence of intraosseous (tissue) edema may be very helpful as would be the presence or absence of a soft tissue mass. For example, a mass would be much more likely to be present in a malignant bone tumor.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here