CT-guided needle biopsies of bone and soft tissue tumors: a pathologist’s perspective

Because bone and soft tissue tumors are not as common as those of other organ systems, few pathologists develop the expertise to diagnose these difficult lesions. Even for experienced bone pathologists, some bone tumors are difficult to diagnose even with large tissue samples. Therefore, core needle biopsies of bone and joint lesions should only be performed at institutions which are supported by experienced bone and soft tissue pathologists. Indeed, the published studies of this procedure were performed at institutions which had a musculoskeletal team consisting of bone radiologists and pathologists. In these studies, accuracy rates ranged from 66–97% [1‐3]. Presumably, in these studies patients were selected on the basis of a likely high yield.

Although a radiologist can almost always hit the lesion with CT guidance, a pathologic diagnosis cannot always be rendered. Necrotic areas and reactive changes are very common in bone and joint lesions, and these areas may be sampled instead of diagnostic areas. Where possible, the radiologist should take at least four cores from different zones of the lesion. Obtaining multiple cores from small lesions is especially difficult. An important concept to remember is that the absence of neoplastic cells in a core biopsy sample does not mean the patient does not have a neoplasm. If radiographs suggest a neoplasm and a core biopsy has no neoplastic cells, a more extensive biopsy is necessary.

Needle samples of radiodense lesions have a much lower probability of yielding diagnostic results (personal experience). In fact, the yield is so low that open biopsy of radiodense lesions is almost always preferable. Radiodense lesions are difficult to penetrate with needles. In addition, there is often considerable crush artifact and the production of bone dust which obscures diagnostic cells. Moreover, the radiodensity is usually due to the proliferation of reactive native bone which can also obscure the underlined lesion. Radiodense lesions in the spine are particularly low yield because the mechanics of the technique makes multiple cores almost impossible to obtain.

Diagnostic material is more likely obtained in those lesions which have very characteristic radiographic appearances. For example, giant cell tumors, Ewing’s sarcomas and many conventional osteosarcomas have such characteristic imaging appearances that a biopsy need only confirm the imaging impression. Only exceptionally does a biopsy provide an unsuspected diagnosis. For example, a radiographically conventional osteosarcoma can be confirmed with only a few malignant cells either from a needle core or an FNA. The presence of osteoid in the biopsy is not necessary if the radiograph is typical and has bone matrix. As an aside, in my consultation practice, I continue to receive needle core samples from lesions such as nonossifying fibroma and fibrous dysplasia, lesions that may be diagnosed by imaging alone. This seems to me to reflect interpretative uncertainty (a disproportionately broad differential diagnosis) and indicates that a biopsy is done with readiness because a lesion is accessible.

The lesions with the highest yield on needle core biopsy are lytic lesions in older patients. Most of these lesions are either metastatic carcinoma or multiple myeloma. These lesions can easily be diagnosed on needle core provided that multiple samples are obtained and that the pathologist performs special stains in equivocal cases.

Lesions in children and young adults have a higher likelihood of being primary bone tumors. These bone tumors may be extremely difficult to diagnose on a needle core. This is because primary bone tumors are often variegated histologically. Areas of necrosis, aneurysmal bone cyst, fibrosis, and reactive bone often obscure an underlying lesion. As a result the diagnostic pathologic material may be present in only a small area of the radiographic lesion. These lesions are even difficult to diagnose with large open biopsy samples. Primary bone tumors are especially difficult to diagnose with needle cores in those lesions where the radiographic features are not characteristic of a specific lesion.

Cartilage lesions are notoriously unreliable for characterization with needle core biopsies. The distinction between low-grade chondrosarcomas and enchondromas, for example, cannot be made on a needle biopsy. Needle biopsies should not be performed in cartilage lesions in order to distinguish benign or low-grade malignant. Distinguishing benign from malignant cartilage tumors is multi-factorial and requires a variety of inputs such as the presence of clinical pain, the overall radiographic interpretation, and the progress of the lesion over time. Moreover, some benign cartilage lesions may have extreme cytologic atypia and may be misdiagnosed as chondrosarcomas. This is especially true with synovial chondromatosis and chondromyxoid fibromas, lesions which commonly show cytologic atypia. Therefore, the diagnosis of low-grade chondrosarcoma based on a needle core should be accepted with great skepticism. Surface cartilage lesions are also not distinguishable with small core biopsies.

Soft tissue lesions can also be problematic. Differentiating benign from malignant based on imaging is not as reliable with soft tissue tumors as it is with bone lesions. All the more reason the tissue sample must be diagnostic. Like bone tumors, however, soft tissue tumors may be highly variegated. Highly cellular areas may alternate with pauci-cellular areas. Therefore, low-grade malignant soft tissue lesions should be diagnosed with great caution. Next to a biopsied low-grade area there may be an unsampled high-grade area. As in bone lesions, the radiologist should sample many different areas. As with bone tumors, soft tissue lesion may show tissue heterogeneity. Therefore, the absence of malignant cells in the needle core does not mean that the lesion itself is unequivocally benign.

These points indicate that the planning of a needle core biopsy should ideally be a multi-disciplinary process with consultation between an orthopedic oncologist, bone and soft tissue pathologist, and involved radiologist prior to biopsy. In some instances, the pathologist may advise the radiologist that the odds of getting a microscopic diagnosis in a particular case are extremely low. In these cases, open biopsy should be the procedure of choice. This is preferable to several attempts at needle core biopsies without diagnostic yield. Likewise, the radiologist should have input from the surgeons who may be called on to do resections based on the biopsy. The needle core tract may need to be included in the resected specimen. Therefore, the direction of the biopsy should be planned with the treating surgeon.