General Principles of Tumors

General approach to musculoskeletal neoplasms

An adequate history and physical examination are the first steps in evaluating a patient with a musculoskeletal tumor. Patients may present to the orthopaedic oncologist with pain, a mass, or an abnormal radiographic finding detected during the evaluation of an unrelated problem. Patients with bone tumors most frequently present with pain. The pain initially may be activity related, but a patient with a malignancy of bone often complains of progressive pain at rest and at night. Patients with benign bone tumors also may have activity-related pain if the lesion is large enough to weaken the bone. Other benign lesions, most notably osteoid osteoma, may cause night pain initially. Conversely, patients with soft-tissue tumors rarely complain of pain but more often complain of a mass. Exceptions to this rule are patients with nerve sheath tumors who have pain or neurologic signs.

Although some tumors show a sex predilection (e.g., female predominance with giant cell tumors), this is rarely of diagnostic significance. Race likewise is of little significance, with the exception that Ewing sarcoma is exceedingly rare in individuals of African descent. Family history occasionally can be helpful, as in cases of multiple hereditary exostosis (autosomal dominant inheritance) and neurofibromatosis (autosomal dominant inheritance). Age may be the most important information obtained in the history, however, because most benign and malignant musculoskeletal neoplasms occur within specific age ranges.

Metastases of unknown origin

In a patient older than age 40 with a new, painful bone lesion, multiple myeloma and metastatic carcinoma are the most likely diagnoses even if the patient has no known history of carcinoma. Prostate cancer and breast cancer are the two most common primary sources for bone metastases. If a patient has no known primary tumor, however, the most likely sources are lung cancer and renal cell carcinoma. Rougraff et al. described the proper evaluation of a patient with suspected metastases of unknown origin. The evaluation begins with a history focusing on any previous malignancies, even in the remote past, followed by a physical examination that includes not only the involved extremity but also the thyroid, lungs, abdomen, prostate in men, and breasts in women. Laboratory analysis should include complete blood cell count, erythrocyte sedimentation rate, electrolytes, liver enzymes, alkaline phosphatase, serum protein electrophoresis, and possibly prostate-specific antigen. Plain radiographs of the involved bone and the chest should be obtained. A whole-body bone scan should be ordered to evaluate other possible areas of skeletal involvement, and a CT scan of the chest, abdomen, and pelvis should be obtained ( Fig. 24.1 ). A mammogram is not routinely indicated as an initial procedure because breast cancer is a rare source of metastases without a known primary lesion. The authors were able to identify the primary lesion in 85% of patients with skeletal metastases of unknown origin using this simple approach. They listed six reasons why the biopsy should not be done until the evaluation is complete: (1) The lesion may be a primary sarcoma of bone that may require a biopsy technique that allows for future limb salvage surgery; (2) another, more accessible lesion may be found; (3) if renal cell carcinoma is considered likely, the surgeon may wish to consider preoperative embolization to avoid excessive bleeding; (4) if the diagnosis of multiple myeloma is made by laboratory studies, an unnecessary biopsy can be avoided; (5) the pathologic diagnosis is more accurate if aided by appropriate imaging studies; and (6) the pathologist and surgeon may be more assured of a diagnosis of metastasis made on frozen section analysis if supported by the preoperative evaluation. This is important if stabilization of an impending fracture is planned for the same procedure.

Staging

Enneking and others have shown the desirability of staging benign and malignant musculoskeletal tumors to aid in treatment decision making, provide some determination of prognosis, and allow meaningful comparisons of treatment methods. Benign and malignant tumors of bone and soft tissue can be staged according to the Enneking staging system ( Table 24.1 ). The stages of benign tumors are designated by Arabic numbers, and malignant tumors are designated by Roman numerals.

Benign tumors are staged as follows: stage 1, latent; stage 2, active; and stage 3, aggressive. Stage 1 lesions are intracapsular, usually asymptomatic, and frequently incidental findings. Radiographic features include a well-defined margin with a thick rim of reactive bone. There is no cortical destruction or expansion. These lesions do not require treatment because they do not compromise the strength of the bone and usually resolve spontaneously. An example is a small asymptomatic nonossifying fibroma discovered incidentally on radiographs taken to evaluate an unrelated injury ( Fig. 24.2 ). Stage 2 lesions also are intracapsular but are actively growing and can cause symptoms or lead to pathologic fracture. They have well-defined margins on radiographs but may expand and thin the cortex. Usually they have only a thin rim of reactive bone. Treatment usually consists of extended curettage ( Fig. 24.3 ). Stage 3 lesions are extracapsular. Their aggressive nature is apparent clinically and radiographically. They do not respect natural anatomic barriers and usually have broken through the reactive bone and possibly the cortex ( Fig. 24.4 ). MRI may show a soft-tissue mass, and metastases may be present in 1% to 5% of patients with these lesions (e.g., giant cell tumor). Treatment consists of extended curettage, marginal resection, or possibly wide resection, and local recurrences are common. Reconstruction may sometimes prove difficult. Some interobserver discrepancy may be present when trying to assign a bone lesion to a particular stage.

Musculoskeletal sarcomas also can be staged according to the surgical staging system as described by Enneking et al. This system was designed to incorporate the most significant prognostic factors into a system of progressive stages that helps to guide surgical and adjuvant treatments. The system is based on the histologic grade of the tumor, its local extent, and the presence or absence of metastases. Low-grade lesions are designated as stage I. These lesions are well-differentiated, have few mitoses, and exhibit only moderate cytologic atypia. The risk for metastases is low (<25%). High-grade lesions are designated as stage II. They are poorly differentiated with a high mitotic rate and a high cell-to-matrix ratio. Stage I and II lesions are subdivided according to the extent of local growth. Stage IA and IIA lesions are contained within well-defined anatomic compartments ( Fig. 24.5 ). Anatomic compartments are determined by the natural anatomic barriers to tumor growth, such as cortical bone, articular cartilage, fascial septa, or joint capsules. Stage IB and IIB lesions extend beyond the compartment of origin ( Fig. 24.6 ). Stage III refers to any lesion that has metastasized regardless of the size or grade of the primary tumor. No distinction is made between lymph node metastases or distant metastases because both circumstances are associated with a poor prognosis.

Alternatively, many orthopaedic oncologists stage musculoskeletal malignancies according to the American Joint Committee on Cancer (AJCC) system. The AJCC staging system for soft-tissue sarcomas ( Table 24.2 ) is based on prognostic variables, including tumor grade (low or high), size (≤5 cm or >5 cm in greatest dimension), depth (superficial or deep to the fascia), and presence of metastases. Stage I tumors are low grade regardless of size or depth. Stage II tumors are high grade; they may be small and any depth or large and superficial. Stage III tumors are high grade, large, and deep. Stage IV tumors are tumors associated with metastases (including local lymph nodes) regardless of grade, size, or depth.

The AJCC system for bone sarcomas ( Table 24.3 ) is based on tumor grade, size, and presence and location of metastases. Stage I tumors, which are low grade, and stage II tumors, which are high grade, are subdivided based on tumor size. Stage I-A and II-A tumors are 8 cm or less in their greatest linear measurement; stage I-B and II-B tumors are larger than 8 cm. Stage III tumors have “skip metastases,” which are defined as discontinuous lesions within the same bone. Stage IV-A involves pulmonary metastases, whereas stage IV-B involves nonpulmonary metastases. The subdivision of stage IV was made because it has been shown that patients with nonpulmonary metastases from osteosarcoma and Ewing sarcoma have worse prognoses than patients with only pulmonary metastases.

Biopsy

In 1982, Mankin et al. reported 18.2% major errors in diagnosis, 10.3% nonrepresentative or technically poor biopsy specimens, and 17.3% wound complications associated with biopsy of musculoskeletal sarcomas. As a result of these complications, the optimal treatment plan had to be altered in 18.2%, including unnecessary amputations in 4.5%. These complications occurred three to five times more frequently when the biopsy was done by a surgeon at a referring institution, rather than by a member of the Musculoskeletal Tumor Society. A series of recommendations were made regarding the technical aspects of the biopsy, stating that whenever possible a patient with a suspected primary musculoskeletal malignancy should be referred before biopsy to the institution where definitive treatment will take place. The study was repeated 10 years later, and the results were essentially unchanged.

A biopsy should be planned as carefully as the definitive procedure. Biopsy should be done only after clinical, laboratory, and radiographic examinations are complete. As stated previously, completion of the evaluation before biopsy aids in planning the placement of the biopsy incision, helps provide more information leading to a more accurate pathologic diagnosis, and avoids artifacts on imaging studies. If the results of the evaluation suggest that a primary malignancy is in the differential diagnosis, the patient should be referred to a musculoskeletal oncologist before biopsy.

Regardless of whether a needle biopsy or an open biopsy is done, the biopsy track should be considered contaminated with tumor cells. Placement of the biopsy is a crucial decision because the biopsy track needs to be excised en bloc with the tumor. The surgeon performing the biopsy should be familiar with incisions for limb salvage surgery and standard and nonstandard amputation flaps. If a tourniquet is used, the limb can be elevated before inflation but should not be exsanguinated by compression to prevent “squeezing” the tumor’s cells into the systemic circulation. Care should be taken to contaminate as little tissue as possible. Transverse incisions should be avoided because they are extremely difficult or impossible to excise with the specimen ( Fig. 24.7 ). The deep incision should go through a single muscle compartment rather than contaminating an intermuscular plane. Major neurovascular structures should be avoided. Soft-tissue extension of a bone lesion should be sampled because this leading edge contains the most viable tumor for making the diagnosis. Care should be taken, however, to sample more than just the pseudocapsule surrounding the lesion. A frozen section should be sent intraoperatively to ensure that diagnostic tissue has been obtained. If a hole must be made in the bone, it should be round or oval to minimize stress concentration and prevent a subsequent fracture, which could preclude limb salvage surgery ( Fig. 24.8 ). The hole should be plugged with methacrylate to limit hematoma formation. Only the minimal amount of methacrylate needed to plug the hole should be used because excessive amounts push the tumor up and down the bone. If a tourniquet has been used, it should be deflated and meticulous hemostasis ensured before closure, because a hematoma would be contaminated with tumor cells. If a drain is used, it should exit in line with the incision so that the drain track also can be easily excised en bloc with the tumor. The wound should be closed tightly in layers. Wide retention sutures should not be used.

A biopsy can be done by fine-needle aspiration, core needle biopsy, or an open incisional procedure ( Table 24.4 ). Most musculoskeletal neoplasms can be diagnosed with a well-done needle biopsy. Fine-needle aspiration may be 90% accurate at determining malignancy; however, its accuracy at determining specific tumor type is much lower because only cells rather than tissue architecture are evaluated. This technique may be best applied when there is a high probability that the diagnosis is known such as metastases or infection and when evaluating lymph nodes. An experienced pathologist is helpful in determining the diagnosis because of the limited sample size obtained. A core needle biopsy uses a larger-gauge needle than a fine-needle aspiration, providing for tissue and preservation of the tissue architecture. The limited amount of tissue obtained may not be adequate, however, for accurate grading or for any additional studies that may dictate subsequent treatment. The few dedicated series that have analyzed outpatient core needle biopsies have reported an overall diagnostic accuracy ranging from 84% to 98%. A study of 252 outpatient core needle biopsies of malignant bone and soft-tissue neoplasms reported an accuracy rate of 97% for determining whether or not a lesion is malignant; core needle biopsy was accurate for a specific histopathologic diagnosis and grade in 81%.

Open biopsy is the gold standard for biopsy of bone and soft-tissue tumors, but complications are greater with incisional biopsy when compared with needle biopsy (e.g., bleeding, infection, tissue contamination). However, this procedure is least likely to be associated with a sampling error, and it provides the most tissue for additional diagnostic studies, such as cytogenetics and flow cytometry. If the administration of chemotherapy is anticipated before further surgery, a central venous access catheter may be placed at the same setting as the biopsy if the frozen section is confirmatory. The definitive procedure can be done immediately after biopsy only if the frozen section diagnosis confirms the clinical and radiographic diagnosis. In cases of discrepancy or doubt, the definitive procedure should be delayed until a firm diagnosis is established. If a giant cell tumor is suspected on clinical and radiographic grounds, definitive curettage can proceed immediately after confirmation of the diagnosis on frozen section. Likewise, if the suspicion of an impending fracture from metastatic carcinoma is confirmed on frozen section, prophylactic fixation can be applied immediately. Conversely, if the frozen section in either of these scenarios exhibited any atypical cells that might represent a sarcoma, definitive surgery should be delayed until the final pathologic evaluation is complete.

Rarely, a primary resection (i.e., excisional biopsy) should be done instead of a biopsy. A small (<3 cm) subcutaneous mass that is unlikely to be malignant may be marginally resected primarily. In the rare circumstance that the lesion turns out to be malignant, the tumor bed can be reexcised with wide margins without adversely affecting the outcome. Primary resection should not be done on larger soft-tissue lesions or lesions deep to the fascia unless the MRI appearance is diagnostic of a benign lesion, such as a lipoma. Some benign bone lesions, such as osteoid osteoma and osteochondroma, have a characteristic radiographic appearance and can be primarily resected, if indicated, without biopsy. A final relative indication for primary resection is a painful lesion in an expendable bone, such as the proximal fibula or distal ulna. If a symptomatic lesion in one of these locations is confined within the cortex and would be resected regardless of a benign or malignant tissue diagnosis, it can be resected without biopsy ( Fig. 24.9 ).

Radiation therapy

Radiation causes cell death by inducing the formation of intracellular free radicals that subsequently cause DNA damage. The sensitivity of a cell to radiation depends on several factors, including (1) the cell’s position in the cell cycle (actively mitotic cells are most sensitive), (2) tissue oxygenation (local hypoxia provides a protective effect because oxygen-free radicals cannot be formed in hypoxic tissue), and (3) the cell’s ability to repair DNA damage or its inability to undergo apoptosis (programmed cell death) in response to this damage.

The dose of radiation is measured in Gray (Gy): 1 Gy is equal to 1 joule of absorbed energy per kilogram; 1 rad is equal to 1 centigray (cGy). The goal of radiation treatment is to deliver the highest possible dose of radiation to the tumor cells while minimizing toxicity to normal tissues. This is accomplished by using linear accelerators that deliver a high dose to the target tissues with sharp lateral field edges that limit the dose to nontarget tissues. Therapeutic advantage also is gained by fractionation of the dose. After a single treatment of 200 cGy, all cells in the most sensitive phase of the cell cycle are killed. Delivering another dose at a specified interval allows additional cells to enter this phase of the cell cycle. In addition, with progressive tumor cell death, previously hypoxic areas of the tumor may become reoxygenated and may become more sensitive to radiation. The interval also allows time for normal cells to repair damage. Most radiation treatment protocols deliver 150 to 200 cGy/day until the target dose is achieved. This dose ranges from 30 to 40 Gy for myeloma to 60 Gy for treatment of a soft-tissue sarcoma.

Most primary bone malignancies are relatively radioresistant. Exceptions are the small blue cell tumors, including multiple myeloma, lymphoma, and Ewing sarcoma, which are each exquisitely sensitive. Carcinomas metastatic to bone, with the exception of renal cell carcinoma, also frequently are sensitive to radiation treatment. For most other bone tumors, radiation has a limited role because local control is achieved better with surgery. Advances in spinal surgery have diminished the frequency of use of radiotherapy for tumors that were previously surgically inaccessible. Radiation therapy can be used to reduce the incidence of local recurrence of malignant soft-tissue tumors treated with marginal resection when the alternative would be a more mutilating resection or amputation. Radiation also can be used for preoperative treatment of soft-tissue sarcomas in the hopes of reducing the tumor volume and making the resection easier.

Radiation therapy is associated with significant acute and long-term complications. Acutely, the most common complication is skin irritation. Initial erythema may progress later to desquamation, especially in patients who also are being treated with cytotoxic drugs. Other common acute side effects include gastrointestinal upset, urinary frequency, fatigue, anorexia, and extremity edema. Late effects include chronic edema, fibrosis, osteonecrosis, and pathologic fracture. Malignant transformation of irradiated tissues (i.e., radiation sarcoma) is being reported with increasing frequency in survivors of childhood and adolescent cancers. These secondary sarcomas occur with a mean lag time of approximately 10 years and often are associated with a poor prognosis. Radiation-induced pathologic fractures also are becoming more common and can be extremely difficult to treat. In a study by Lin et al., the incidence of pathologic fracture was 29% at 5 years after treatment of a soft-tissue sarcoma of the thigh if treatment included radiation therapy and wide resection with periosteal stripping. This risk increased to 47% in female patients and to 66% in female patients who received chemotherapy. Radiation therapy in children has several adverse sequelae, such as scoliosis, kyphosis, chest wall deformities, hypoplasia of the ilium, and limb-length discrepancy, as a result of radiation-induced growth arrest. Radiotherapy is rarely used for benign conditions; possible exceptions include an extensive pigmented villonodular synovitis that cannot be controlled by surgery or a large spinal giant cell tumor.

In addition to conventional external beam radiation, radiation can be delivered by brachytherapy (from the Greek, brachys, meaning “close”). By this method, hollow catheters are implanted in the tumor bed at the time of resection ( Fig. 24.10 ). These catheters exit through the skin. Postoperative radiographic evaluation and computer calculations determine the optimal loading of the catheters with radioisotopes. This technique allows for high doses to be delivered to the target tissues. The radiation levels fall off rapidly at the edges of the field, sparing normal tissues.

Chemotherapy

Before the routine use of chemotherapy for osteosarcoma, patients usually were treated with immediate wide or radical amputation on diagnosis. This approach usually treated the local disease adequately. Nevertheless, 80% of patients eventually died of metastatic disease even if metastasis was not evident at presentation. From this, it can be deduced that 80% of patients with apparently localized osteosarcoma actually have undetectable metastases, or micrometastases, on presentation. With the use of modern chemotherapy protocols, the current 5-year survival rate for osteosarcoma is approximately 70%. Similar numbers are available regarding the treatment of Ewing sarcoma. Similarly, chemotherapy has a well-defined role in the treatment of other high-grade malignancies of bone, such as malignant fibrous histiocytoma, and high-grade soft-tissue malignancies of childhood, such as rhabdomyosarcoma. The role of chemotherapy is less well defined for adult soft-tissue malignancies, with most investigations showing modest improvements in outcome. In general, chemotherapy is not useful for cartilaginous lesions and most low-grade malignancies.

Adjuvant chemotherapy refers to chemotherapy administered postoperatively to treat presumed micrometastases. Neoadjuvant chemotherapy refers to chemotherapy administered before surgical resection of the primary tumor. No study has proved a survival advantage with regard to the timing of chemotherapy; however, multiple authors have cited several theoretical advantages of neoadjuvant chemotherapy over adjuvant chemotherapy. Preoperative chemotherapy frequently causes regression of the primary tumor, making a successful limb salvage operation easier. In a study by Malawar et al., 9 of 12 lesions that initially were deemed unresectable were treated later with limb salvage surgery after chemotherapy-induced tumor regression. Neoadjuvant chemotherapy followed by surgical resection allows for histologic evaluation of the effectiveness of treatment. This is one of the most valuable prognostic indicators (i.e., percent tumor necrosis) of successful long-term outcome. In addition, histologic evaluation may lead to alteration of further chemotherapy in poor responders. Preoperative chemotherapy theoretically may decrease the spread of tumor cells at the time of surgery, and neoadjuvant chemotherapy usually can be started immediately, effectively treating micrometastases at the earliest time possible and avoiding tumor progression, which may occur during any delay before surgery. This allows time to plan the operation properly, including the possible manufacturing of a custom implant. It also allows time for the patient and the family to consider fully the options of limb salvage surgery versus amputation. It has been suggested, however, that neoadjuvant chemotherapy may increase significantly the risks of perioperative complications, especially delayed wound healing and infection. Others have not found this to be true, and currently most orthopaedic oncologists favor preoperative chemotherapy with the definitive procedure performed 3 to 4 weeks after the last dose has been administered. Chemotherapy is restarted 2 weeks postoperatively if the wound has healed.

Although it is presumed that most malignancies arise from a single cell, the actual tumor is composed of a heterogeneous population of cells. This is the result of rapid turnover and genetic lability of these cells. As a result, various cells within the same tumor evolve different mechanisms of chemoresistance. To combat this diversity in resistance, most chemotherapy regimens involve combinations of toxic drugs. Table 24.5 lists the most common agents used in the treatment of bone and soft-tissue sarcomas and their most common dose-limiting side effects. This field is rapidly changing, but certain general statements can be made. These drugs are most effective when the tumor against which they are directed is small. Combinations of these drugs are more effective than single agents. Dosage, sequence of drugs, and schedule seem to be important in achieving the maximal response. All have toxicity for normal tissues and should be given in a controlled setting by someone skilled in their use.

Amputation versus limb salvage

Advances in diagnostic imaging, chemotherapy, radiation therapy, and surgical technique for resection and reconstruction now allow limb salvage to be a reasonable option for most patients with bone or soft-tissue sarcomas. Specifically, preoperative radiation therapy for soft-tissue sarcomas and neoadjuvant chemotherapy for bone sarcomas have helped surgeons to successfully resect some tumors that in the past would have been deemed unresectable. Rarely, involvement of neurovascular structures, a displaced pathologic fracture, or complications secondary to a poorly performed biopsy preclude the possibility of limb salvage. More often, however, the choice between limb salvage and amputation must be made on the basis of the expectations and desires of the individual patient and the family. Simon described four issues that must be considered whenever contemplating limb salvage instead of an amputation:

• 1.

  Would survival be affected by the treatment choice?

• 2.

  How do the short-term and long-term morbidity compare?

• 3.

  How would the function of a salvaged limb compare with that of a prosthesis?

• 4.

  Are there any psychosocial consequences?

Several studies have shown the effect of treatment choice on survival in patients with osteosarcoma. With the use of multimodal treatment including surgery and chemotherapy, long-term survival for patients with osteosarcoma has improved from approximately 20% to approximately 70% in most series. For osteosarcoma of the distal femur, the rate of local recurrence after wide resection and limb salvage is 5% to 10%. This is equivalent to the rate of local recurrence after a transfemoral amputation in this setting. Although hip disarticulation is associated with an extremely low rate of local recurrence, no study has shown a survival advantage for this technique. Although local recurrence is associated with an extremely poor prognosis, no study has proved any one of these surgical techniques (i.e., limb salvage, transfemoral amputation, hip disarticulation) to be superior in terms of survival, provided that wide surgical margins are obtained. From these data it can be hypothesized that patients with a local recurrence despite wide margins may represent a subset of patients with especially aggressive disease or chemotherapy-resistant disease who would do poorly regardless of the surgical procedure. With regard to overall patient survival, the most important technical aspect of the surgical procedure is the attainment of a wide margin regardless of whether this is achieved by amputation or local resection.

Amputation for malignancy can be technically demanding, often requiring nonstandard flaps for closure or bone graft augmentation to obtain a more functional residual limb. Complications include infection, wound dehiscence, a chronically painful limb, phantom limb pain, and appositional bone overgrowth requiring revision surgery. Limb salvage is associated, however, with greater perioperative and long-term morbidity compared with amputation. Limb salvage requires a much more extensive surgical procedure with greater risks for infection, wound dehiscence, flap necrosis, blood loss, and deep venous thrombosis. Long-term complications of limb salvage vary depending on the type of reconstruction. These include periprosthetic fractures, prosthetic loosening or dislocation, nonunion of the graft-host junction, allograft fracture, leg-length discrepancy, and late infection. A patient with a salvaged limb is much more likely to need multiple future operations for treatment of complications. After initially successful limb salvage surgery, one third of long-term survivors ultimately may require an amputation depending on the location of the tumor and the type of reconstruction.

With regard to function, location of the tumor is the most important issue. Resection of an upper extremity lesion with limb salvage, even with the sacrifice of one or two major nerves, generally provides better function than amputation and subsequent prosthetic fitting. Similarly, resection of a proximal femoral or pelvic lesion with local reconstruction generally provides better function than would be possible after a hip disarticulation or hemipelvectomy. Around the ankle and foot, however, large sarcomas frequently are treated with amputation followed by prosthetic fitting. Treatment of sarcomas around the knee must be individualized.

Most patients with osteosarcomas around the knee are treated with wide resection with prosthetic knee replacement or transfemoral amputation. Less commonly performed operations include osteoarticular allograft reconstruction, allograft arthrodesis, and rotationplasty. In a study of patients with osteosarcoma, Otis, Lane, and Kroll showed that, compared with transfemoral amputees, patients who had undergone resection and prosthetic knee replacement showed higher self-selected walking velocities and a more efficient gait with regard to oxygen consumption. Many of the transfemoral amputees were functioning at greater than 50% of their maximal aerobic capacity at free walking speeds. At greater than 50% of the maximal aerobic capacity, anaerobic mechanisms are required to sustain muscle metabolism, and endurance is greatly decreased. The problem is compounded by the decreased cardiac function in many of these patients as a result of doxorubicin-induced cardiomyopathy.

Harris et al. compared the long-term function of amputation, arthrodesis, and arthroplasty for tumors around the knee. They showed that patients who had amputations had difficulty walking on steep, rough, or slippery surfaces but were active and were the least worried about damaging the affected limb. Patients with an arthrodesis performed the most demanding physical work and recreational activities but had difficulty with sitting, especially in the back seats of cars, theaters, or sports arenas. Patients who had an arthroplasty generally led more sedentary lives and were more protective of the limb but had little difficulty with activities of daily living. These patients also were the least self-conscious about the limb.

In patients who are long-term survivors after resection of an extremity sarcoma, the probability of limb survival is associated with the type of reconstruction and the location of the tumor. A successful arthrodesis is more durable in the long term than a mobile joint reconstruction. Regarding prosthetic or allograft-prosthetic composite reconstructions, location is the most important issue, with proximal reconstructions generally outlasting more distal reconstructions. (This is the inverse of the prognosis for overall patient survival, with distal sarcomas having a better prognosis than proximal sarcomas.) Proximal femoral reconstructions generally outlast distal femoral reconstructions, which generally outlast proximal tibial reconstructions.

No study has shown a significant difference between amputation and limb salvage with regard to psychologic outcome or quality of life in long-term survivors of sarcoma. The choice of limb salvage or amputation involves more than the question of whether the lesion can be resected with wide margins. The patient ultimately must make the final decision in light of long-term goals and lifestyle decisions.

Margins

When describing an oncologic surgical procedure, it is imperative that the surgical margin be appropriately defined. The terms amputation and resection mean little without a modifier describing the margin. This is especially important when evaluating surgical procedures and outcomes in the literature. In orthopaedic oncology, the surgical margin is described by one of four terms: intralesional, marginal, wide, or radical. Amputations and limb-sparing resections may be associated with any of the four types of margins, and the margin must be specifically defined with each procedure ( Figs. 24.11 and 24.12 ).

An intralesional margin is one in which the plane of surgical dissection is within the tumor. This type of procedure is often described as “debulking” because it leaves behind gross residual tumor. This procedure may be appropriate for symptomatic benign lesions when the only surgical alternative would be to sacrifice important anatomic structures. This also may be appropriate as a palliative procedure in the setting of metastatic disease.

As musculoskeletal tumors grow, they compress the surrounding tissues and appear to become encapsulated. This surrounding reactive tissue often is referred to as the pseudocapsule. A marginal margin is achieved when the closest plane of dissection passes through the pseudocapsule. This type of margin usually is adequate to treat most benign lesions and some low-grade malignancies. In high-grade malignancy, however, the pseudocapsule often contains microscopic foci of disease, or “satellite” lesions. A marginal resection often leaves behind microscopic disease that may lead to local recurrence if the remaining tumor cells do not respond to adjuvant chemotherapy or radiation therapy. Despite an increased risk of local recurrence, a marginal resection may be preferable if the alternative is a more mutilating procedure. Improvements in preoperative radiation therapy and neoadjuvant chemotherapy have made marginal resections an acceptable alternative to amputation in some selective circumstances.

Wide margins are achieved when the plane of dissection is in normal tissue. Although no specific distance is defined, the entire tumor remains completely surrounded by a cuff of normal tissue. The quality of a margin is more important than the quantity (thickness) of the margin. For instance, a fascial margin provides a better plane for containing tumor spread than does a similar or thicker plane of subcutaneous tissue. If the plane of dissection touches the pseudocapsule at any point, the margin should be defined as being marginal and not wide. Although sometimes impossible to achieve, wide margins are the goal of most procedures for high-grade malignancies.

Radical margins are achieved when all the compartments that contain tumor are removed en bloc. For deep soft-tissue tumors, this involves removing the entire compartment (or multiple compartments) of any involved muscles. For bone tumors, this involves removing the entire bone and the compartments of any involved muscles. Radical operations were previously the procedures of choice for most high-grade neoplasms; however, with improvements in imaging studies, radical procedures now are rarely performed because equivalent oncologic results usually can be obtained with wide margins.

From an oncologic standpoint, there are eight different surgical procedures because resections and amputations may be defined further by any of the four margins. Amputations usually achieve wide margins (e.g., a high transfemoral amputation for an osteosarcoma of the distal femur) or radical margins (e.g., a hip disarticulation for a femoral lesion), but this is not always the case. A hemipelvectomy for a large intrapelvic tumor may allow only marginal margins to be obtained and would be referred to as a marginal amputation. Rarely, a palliative procedure or an inappropriate amputation level leaves behind gross residual disease. These procedures would be referred to as intralesional amputations. Likewise, limb-sparing resections of bone or soft-tissue tumors can be categorized by any one of the types of margins, although radical resections of bone tumors are extremely rare.

Curettage

Many benign bone tumors are treated adequately by curettage. Compared with resection, curettage is associated with a higher rate of local recurrence; however, curettage often allows for a better functional result. Although this is not a technically difficult procedure, the surgeon should adhere strictly to several principles to avoid an unacceptably high rate of local recurrence, especially with more aggressive benign tumors.

Curettage is done by first making a large cortical window over the lesion. This window must be at least as large as the lesion itself. If the window is smaller than the lesion, the surgeon inevitably leaves residual tumor on the undersurface of the near cortex. The bulk of the tumor is scooped out with large curets. Next, the cavity is enlarged back to normal host bone in each direction with a power burr. (Use of a power burr is mandatory for curettage of bone tumors.) Finally, the cavity and the wound should be copiously irrigated to remove any debris and tumor cells. These are the minimal requirements for a “simple” curettage.

“Extended” curettage includes the use of adjuvants, such as liquid nitrogen, phenol, polymethyl methacrylate, or thermal cautery ( Fig. 24.13 ) to extend destruction of tumor cells. Several authors have reported greatly reduced recurrence rates of aggressive tumors with the use of adjuvants. The recurrence rate after extended curettage for giant cell tumors is now approximately 10%. Although not proved in randomized trials, this seems to be a great improvement compared with the 25% to 50% recurrence rate in historic controls reported before the routine use of adjuvants.

Although each adjuvant treatment has its proponents, no study has proved that any one is superior, with each having advantages and disadvantages. Cryosurgery with liquid nitrogen is effective at extending the tumor kill. Studies have shown it to be superior to phenol and methacrylate at creating a rim of necrotic bone (≤14 mm) around experimental cavities in animal and cadaver models. Liquid nitrogen usually is applied by the “direct pour” technique and may be associated with greater complications, including pathologic fracture and nerve injury. Phenol, conversely, has relatively poor penetration into bone (<1 mm). Although it is relatively easy to use, serious complications have been reported when phenol was inadvertently applied to the surrounding normal tissues. Adjuvant treatment also can be accomplished through thermal cautery, such as with an argon beam coagulator. Studies have shown the depth of necrosis in cancellous bone treated with argon beam coagulation to be approximately 4 mm. We have extensive experience with the use of argon beam coagulation and have noted no complications that can be attributed directly to its use. Finally, despite some disagreement in the literature, polymethyl methacrylate bone cement may act as an adjuvant through its heat of polymerization or through direct toxicity of the monomer. It is easily applied and can be used as a filling agent in conjunction with other adjuvants.

The final issue that must be considered involves filling the cavity left after curettage. Options include autogenous bone graft, allograft, demineralized bone matrix, artificial bone graft substitutes, and bone cement. Autogenous bone graft provides the most rapid and most reliable healing rate because it is osteogenic, osteoinductive, and osteoconductive, but it is associated with additional morbidity at the harvest site, and it may not be available in sufficient quantity to fill a large cavity. Autogenous bone graft must be harvested using a different set of instruments to prevent contamination of the donor site. Even though it is only osteoconductive, cancellous allograft is reliably incorporated. It is readily available in large quantities and does not involve any further operative morbidity. Although allograft is associated with the theoretical risk of disease transmission, we are not aware of any reported cases of hepatitis or human immunodeficiency virus transmission through the use of freeze-dried cancellous allograft. Another alternative is demineralized bone matrix. The material is osteoconductive, but in contrast to cancellous allograft, demineralized bone matrix also is osteoinductive. Artificial bone graft substitutes (e.g., calcium sulfate, calcium phosphate) are osteoconductive, are easy to use, and are readily available. They can be used alone or in combination with autogenous bone graft, bone marrow aspirates, or demineralized bone matrix. Early reports have shown their efficacy with regard to filling relatively large defects. Finally, bone cement can be used as a filling agent. In addition to its use as an adjuvant, it has the advantage of providing immediate stability, which makes rehabilitation easier and lessens the risk of pathologic fracture. Another advantage of bone cement is associated with the detection of local recurrence. Although tumor recurrences are difficult to recognize after a tumor cavity has been filled with bone graft or bone graft substitutes, recurrent tumor is easily recognized as an expanding lucency adjacent to bone cement. One potential disadvantage of bone cement (although not proved) is that it may lead to early joint degeneration secondary to biomechanical alteration of the subchondral bone. Adding a layer of bone graft to the subchondral bone before cement may help minimize the suggested biomechanical alteration. Some authors subsequently have recommended routine removal of the cement at a later date and replacement with bone graft.

We currently use argon beam coagulation as adjuvant treatment after curettage. For most benign lesions, the defects are filled with freeze-dried cancellous allograft chips or with a calcium sulfate/calcium phosphate bone graft substitute. For more aggressive benign lesions, such as giant cell tumors, we usually use bone cement to fill the defect and consider adjuvant fixation if the defect is thought to need additional structural support. We do not routinely remove the cement at a later date to decrease the theoretical risk of degenerative joint disease.

Resection and reconstruction

Currently, most musculoskeletal malignancies are treated with local resection and reconstruction. Aggressive benign neoplasms also can be treated in this manner. The goal of resection of a malignancy is to achieve wide surgical margins if possible. If this is impossible because of anatomic constraints, a marginal resection combined with adjuvant or neoadjuvant treatment (e.g., radiation for a soft-tissue sarcoma) may be preferable to an amputation, although this decision must be made on an individual basis in conjunction with the patient and family. A marginal resection usually is adequate for most benign neoplasms. Specific techniques for resection are discussed later in this chapter.

Although allograft arthrodesis still has a role in some circumstances, most reconstructions involve preserving a mobile joint, for which three general options are available: osteoarticular allograft reconstruction, endoprosthetic reconstruction, and allograft-prosthesis composite reconstruction. (An additional option, rotationplasty, is discussed later in the chapter.) In general, oncologic reconstructions involve higher complication rates than do standard total joint arthroplasties because of the extensive nature of the operation, the extensive tissue loss, and the compromising effects of associated radiation and chemotherapy. In addition, these reconstructions often are done on young patients who are extremely active. Some complications, such as wound necrosis and infection, are universal to all types of reconstructions. Other complications are more specific to the method of reconstruction. Although each method has proponents, we have made the most extensive use of endoprosthetic reconstruction, reserving other methods for specific indications.

Osteoarticular allografts offer several attractive advantages, including the ability to replace ligaments, tendons, and intraarticular structures. Several authors have reported success with this method of reconstruction; however, other authors have reported high rates of complications, including nonunion at the graft-host junction, fatigue fracture, infection, articular collapse, dislocation, degenerative joint disease, and failure of ligament and tendon attachments. Osteoarticular allografts may have a role as a temporary measure to preserve an adjacent physis in an immature patient when the alternatives include amputation or sacrifice of both physes. A proximal tibial osteoarticular allograft could be used in an immature patient in an attempt to preserve the distal femoral physis until skeletal maturity. This could be converted later to an endoprosthetic reconstruction when it becomes necessary.

Allograft-prosthesis composites may provide a long-term solution for some patients. They avoid the complications of degenerative joint disease and articular collapse while still preserving the ability to attach soft-tissue structures directly, such as the patellar tendon or the hip abductors. They are associated, however, with fatigue fracture, infection, and nonunion at the graft-host junction. Although many surgeons use allograft-prosthesis composite as their primary method of reconstruction, our main indication is an inadequate length of remaining host bone to secure the stem of an endoprosthesis. We still use a tumor prosthesis for reconstruction with allograft for fixation to the remaining host bone ( Fig. 24.14 ).

Endoprosthetic reconstruction also may provide long-term function for some patients and is associated with its own complications. Endoprosthetic reconstruction provides the advantage of predictable immediate stability that allows for quicker rehabilitation with immediate full weight bearing. Most endoprostheses are modular, allowing for incremental limb lengthening as an immature patient grows. Improvements in implant materials have greatly increased the durability of modern endoprostheses; however, all are associated with long-term complications if a patient is cured of disease. Polyethylene wear is still a limiting factor for articulating surfaces, but the inserts are easily replaceable in most prostheses. Fatigue fracture can occur at the rotating hinge, but this, too, is easily replaceable. Fatigue fracture at the base of the intramedullary stem where it attaches to the body of the prosthesis is more problematic. In this location, extraction of the remaining stem can be extremely difficult.

Segmental bone and joint prostheses are most commonly secured through composite fixation. An intramedullary stem is fixed with cement, and the shoulder region of the prosthesis is constructed with a porous coating with the goal of promoting late extramedullary cortical bridging ( Fig. 24.15 ).

Initial fixation with cement provides immediate stability for quick rehabilitation. The purpose of the extramedullary cortical bridging is to serve as a purse string to protect the cement-bone interface from particulate debris generated at the articulating surface and to provide additional structural support protecting the junction of the base of the stem with the body of the prosthesis. This area is otherwise susceptible to fatigue fracture as a result of stress concentration. Although its benefit has not been established, bone grafting the shoulder region of the prosthesis to promote extracortical bridging has been advocated by several authors.

Considerations for pediatric patients

For pediatric patients, future limb-length inequality must be considered. For patients who are near skeletal maturity, the reconstructed limb can be lengthened 1 cm at the initial procedure. Also, epiphysiodesis of the contralateral limb can be done at the appropriate age to preserve limb-length equality or minimize inequality. For younger patients, however, other options should be considered. Although amputation and rotationplasty were previously considered the only reasonable treatments for very young patients with bone sarcomas, use of expandable prostheses currently is gaining support.

We have gained considerable experience with use of the Repiphysis Expandable Prosthesis (Wright Medical Technology, Arlington, TN) ( Fig. 24.16 ). The surgical technique for implantation of this device is similar to that of other endoprostheses ( Fig. 24.17 ). The postoperative course, rehabilitation, function, and complications likewise are similar. The device is unique, however, in that it uses energy stored in a compressed spring to allow for future expansion of the prosthesis as the child grows. When a leg-length discrepancy develops, the child is scheduled for an expansion ( Fig. 24.18 ). The procedure is done in the fluoroscopy suite with the patient under light sedation. The locking mechanism on the prosthesis is identified using fluoroscopy, and an electromagnetic coil is placed over the patient’s leg at that level. The electromagnetic coil is activated for 20 seconds, which heats an element in the prosthesis, melting a small segment of polyethylene and allowing controlled expansion of the spring. The leg lengths are reevaluated under fluoroscopy, and the procedure is repeated one or two times as necessary. We have been able to gain 0.5 to 1.5 cm during each scheduled expansion session. Expansion sessions can be scheduled 4 weeks apart if needed to allow the operated leg to “catch up.” After the expansion sessions, patients usually are able to ambulate immediately without an assistive device. Although this device is not as durable and mechanical problems are common, complications are usually relatively easy to address. Moreover, skeletally immature patients treated for bone sarcomas are allowed to maintain limb-length equality at the completion of growth.

Upper extremity

In contrast to the lower extremity, even the best artificial limb fails to provide comparable function in the upper extremity. Modern imaging and surgical techniques allow for limb salvage in most circumstances. Resections of the proximal humerus frequently require sacrifice of the axillary nerve, and resections of the humeral shaft frequently require sacrifice of the radial nerve. Even with sacrifice of three major nerves, limb salvage usually provides better function than an artificial limb. If the median or ulnar nerve must be sacrificed, limb salvage still may be worthwhile if functioning muscles are available for transfers. One indication for amputation is extensive neurovascular involvement. A displaced pathologic fracture may be a relative indication.

Resection of the shoulder girdle

Tumors of the scapula frequently are complicated by extension into the glenohumeral joint, requiring extraarticular resection of the humeral head en bloc with the scapula. Likewise, the biceps tendon provides a passageway for tumors of the proximal humerus to extend into the joint, and resection often requires extraarticular partial scapulectomy. To create a standard terminology for various surgical procedures for shoulder girdle resection and to allow meaningful comparison of results, Malawer et al. proposed a classification of these procedures. They noted that previous concepts have not adequately described surgical margins, the relationship of the tumor to anatomic compartments, the status of the glenohumeral joint, the magnitude of the surgical procedure, or the status of the abductor mechanism.

The proposed system is based solely on the structures removed, reflecting the type of resection and its relationship to the glenohumeral joint, and indicates a progressive increase in the magnitude of the surgical procedure. Additionally, it indicates the status of the abductor mechanism. Procedures are divided into six types: type I, intraarticular proximal humeral resection; type II, partial scapular resection; type III, intraarticular total scapulectomy; type IV, extraarticular total scapulectomy and humeral head resection; type V, extraarticular humeral head resection; and type VI, extraarticular humeral and total scapular resection. Each type is subdivided according to the status of the abductor mechanism: A, intact, or B, partial or complete resection ( Fig. 24.19 ).