Tumors

Grade

The grade of the tumor is a measure of its potential to metastasize. It is based primarily on histologic features and requires a preoperative biopsy. A sarcoma is classified as either low grade or high grade. Generally, a low-grade lesion is less biologically active and requires a relatively conservative surgical procedure. Conversely, a high-grade lesion usually necessitates a more radical procedure because of its more aggressive nature.

Local Extent

Factors related to the local extent of a tumor include its size and degree of involvement of adjacent tissues. Sarcomas tend to grow centrifugally along pathways of least resistance and are contained in part by a pseudocapsule as they extend into adjacent tissues. A malignant lesion may remain confined within the pseudocapsule (intracapsular); generally, however, malignant cells often extend beyond these capsular boundaries. If a lesion extends through its capsule but is still confined within a single anatomic compartment, it is considered extracapsular and intracompartmental. If the tumor extends into an adjacent compartment, it is classified as extracompartmental. Extracompartmental spread may occur via direct tumor invasion of an adjacent compartment or by contamination resulting from fracture, hemorrhage, or an operative procedure such as an unplanned resection or poorly planned biopsy. Generally, lesions with more advanced local extension, including involvement of neurovascular structures or joints, require excision of more adjacent tissue than smaller tumors.

Metastases

The third component of the staging system is the presence or absence of nodal or distant metastases. Determination of metastases is usually accomplished with computed tomography (CT) and radionuclide bone scanning, but more recently, CT with positron emission tomography and whole-body MRI have been advocated as possible alternatives. Regional lymph node involvement is much less common with musculoskeletal sarcomas than are pulmonary metastases, but both are equally poor prognostic factors.

Bone Tumors

MRI is the most sensitive imaging modality for detecting and delineating bone tumors, especially tumors involving the marrow cavity. The MRI appearance of most osseous lesions is very nonspecific, however, and conventional radiographs are essential for evaluating a primary bone tumor. Radiographs should be obtained early in the work-up of a symptomatic patient because they are inexpensive and provide the most specific information of any modality regarding the true nature of a lesion. The radiographic findings and degree of clinical suspicion dictate further work-up. If an aggressive osseous lesion is identified on conventional radiographs, MRI is useful in the preoperative assessment of these patients because it is the best modality for local staging. If a bone lesion is clearly benign radiographically, MRI generally is unnecessary.

For a patient with normal radiographs, a radionuclide bone scan often is the next study obtained, as it can help localize the abnormality and allow for an appropriate field of view on MRI. Once a focal abnormality is detected, MRI is useful for further characterization. Even with a negative bone scan, MRI can detect radiographically occult intramedullary lesions and should be obtained in a patient with a known primary tumor and focal symptoms or laboratory abnormalities that suggest osseous metastases ( Fig. 7.1 ). In this setting, a larger field of view is obtained with a marker placed on the skin at the site of pain in order for the region of interest to be appropriately evaluated.

Soft Tissue Tumors

In a patient with a suspected soft tissue mass, conventional radiographs still should be obtained because they may reveal bone involvement or soft tissue calcifications that might be missed with MRI. In many cases, the MRI appearance of a soft tissue mass is so characteristic that a confident, specific diagnosis can be provided, obviating further work-up. Even if the MRI features do not allow a specific diagnosis to be made, MRI is still useful for staging these lesions.

Important MRI Features ( Box 7.1 )

For osseous and soft tissue lesions, the crucial factors influencing resectability that should be addressed in the MRI report include intraosseous and extraosseous tumor extent and neurovascular, joint, and nodal involvement.

Intraosseous Tumor Extent

Intraosseous tumor extent is best determined with T1-weighted (T1W) and short tau inversion recovery (STIR)/fat-saturated T2-weighted (T2W) imaging ( Fig. 7.2 ); however, the intraosseous extent of tumor may be overestimated with STIR because it can be difficult to separate an intraosseous tumor from peritumoral edema on these images. MRI also is able to detect skip lesions (foci of tumor that are not contiguous with the primary lesion), which may be missed with scintigraphy.

Neurovascular or Joint Involvement

Identification of neurovascular involvement is crucial ( Fig. 7.3 ). Such involvement may preclude the possibility of a limb-sparing procedure because the functional status of a patient with a denervated limb after surgery may be worse than that achieved with an amputation. MRI is highly accurate in showing a lack of neurovascular involvement when a clear plane of normal tissue is shown between nerves or vessels and tumor. Gross tumor invasion usually is easily diagnosed, but if there is equivocal tumor involvement, this should be reported as such. The structures can be reassessed at the time of surgery. On a practical note, an anatomic atlas should be consulted in most cases to determine the expected position of pertinent nerves and vessels. Otherwise, neurovascular involvement might be overlooked if these structures are completely obliterated by a tumor.

Because each joint is a distinct compartment, articular invasion changes the stage of a tumor and should be critically evaluated on every scan. MRI is very accurate for excluding joint involvement when the joint margins appear free of tumor, but is less accurate when the tumor is in close proximity to the joint. Close proximity results in a tendency to overcall joint invasion, which could result in an unnecessarily radical surgical procedure.

Postchemotherapy

Survival of patients with musculoskeletal sarcomas has improved with the development of better adjuvant chemotherapeutic regimens. Assessing the degree of tumor response to chemotherapy is important for establishing the patient’s prognosis and for planning further therapy. If viable tumor cells constitute less than 10% of a lesion after therapy, this indicates a good response (a “responder”), whereas more than 10% represents a poor response (a “nonresponder”). Currently, this response is determined after resection of the tumor, but several series have evaluated the use of MRI in this setting with conflicting results.

Changes in tumor size, signal intensity, or adjacent edema on conventional sequences are not sufficiently predictive to separate responders from nonresponders. Similarly, because tumor and non-neoplastic reactive tissue enhances on standard, postgadolinium T1W images, this technique is also unreliable for this purpose. Dynamic enhancement patterns on gadolinium-enhanced, rapid gradient echo–T1W sequences have shown a high degree of correlation with response or nonresponse because the residual tumor enhances more quickly than the reactive tissue does. We do not use these methods, however, because they are time consuming, technically challenging, and still not reliable enough to replace biopsy and histology.

Postsurgery and Postradiation

MRI is valuable for detecting tumor recurrence after surgical or radiation therapy, primarily because of its superb soft tissue contrast. MRI is sometimes too sensitive in this regard, as postsurgical and postradiation changes in tissues can produce signal intensity that may be mistaken for neoplasm. Careful analysis of T1W, T2W, and STIR images, combined with an understanding of a few basic principles, can markedly improve the diagnostic accuracy of MRI in this setting. Postcontrast imaging may also be beneficial in certain cases, as described subsequently.

A lack of increased signal intensity on T2W or STIR images is a strong predictor of nonrecurrence, as a recurrent tumor usually shows high signal intensity on these images. There are other non-neoplastic causes of increased signal intensity in these patients, however, which can mimic a tumor, including radiation-induced edema and postoperative fluid collections such as hematoma, seroma, or abscess. Certain features help to separate these entities.

Surgery and radiation therapy often result in edema or hemorrhage within tissues, but the absence of a discrete mass is strong evidence against tumor recurrence. This can be evaluated on T1W images by looking for loss of the normal fatty marbling within muscle or distortion of the intermuscular fascial planes. The presence of normal skeletal muscle architecture in these regions on T1W images (normal “texture sign”) is highly predictive of nonrecurrence despite the presence of increased signal intensity on T2W images or enhancement after gadolinium administration ( Fig. 7.4 ).

If a mass is discovered, administration of intravenous (IV) gadolinium may be helpful for further characterization. A postoperative lymphocele, seroma, or abscess appears as a high signal intensity mass on T2W images but does not show internal enhancement on postgadolinium T1W images ( Fig. 7.5 ). If an enhancing mass is identified, biopsy is indicated because recurrent tumor is likely ( Fig. 7.6 ); however, post-therapy granulation tissue also can enhance and produce an identical appearance.