oku10_17

2: System ic D iso rd ers Chapter 17 Musculoskeletal Oncology Kevin B. Jones, MD Basic Principles Population Science Cancer is a common disease, newly affecting more than 1 million Americans each year; however, neoplasms that primarily affect musculoskeletal tissues are rela- tively rare. According to the Surveillance, Epidemiol- ogy, and End Results (SEER) database, an estimated 2,570 individuals were diagnosed with bone sarcomas, with 1,470 deaths in 2008. Soft-tissue sarcomas were more common, with 10,660 diagnoses and 3,820 deaths in 2008.1 Recent reviews of the SEER database have eluci- dated several epidemiologic facts about specific sar- coma subtypes. Osteosarcoma was confirmed to have a bimodal distribution in age of onset, arising as a pri- mary malignancy in adolescents and young adults but as a secondary cancer or complication of Paget disease of bone in elderly patients.2 Osteosarcoma survival rates have not improved over the past 20 years. Ana- tomic location, age, and stage at presentation are each critical for prognosis. Ewing sarcoma was found to be more prevalent in Caucasians than in persons of African or Asian de- scent.3 It also was found that chondrosarcoma survival rates have not improved; tumor grade and stage of dis- ease remain the only independent predictors of sur- vival.4 For soft-tissue sarcomas in general, age, surgical resection, use of radiation, and tumor grade and size each correlated with survival.5 Synovial sarcomas pres- ent across a range of ages but have a better prognosis in younger patients.6 For patients with clear cell sarcomas, nodal as opposed to distant metastasis results in a dra- matically better prognosis.7 Age younger than 16 years and disease that is localized, surgically resectable, and does not involve the lymph nodes are predictive factors of long-term survival in patients with epithelioid sarco- mas.8 Molecular Biology Sarcomas can be grouped into those with abundant cy- togenetic and genetic perturbations and those with bal- anced, reciprocal translocations. The biologic under- standing of representative tumors in each group is progressing rapidly but has produced only minimal im- pact on therapeutic treatments. Osteosarcoma and pleomorphic soft-tissue sarcomas are prototype, complex, genotype sarcomas. Insights into their pathophysiology have arisen from their in- creased incidence in hereditary cancer syndromes such as Li Fraumeni (from p53 disrupting mutations), con- genital retinoblastoma, and Rothmund-Thomson syn- drome (from truncating mutations in the RECQL4 he- licase.) Mouse models of these sarcomas, using targeted disruption of varied tumor suppressor genes, have re- cently been described. Combined disruption of both p53 and pRb in preosteoblasts generated osteosarco- mas that mimic the human disease.9 Disruption of Kras and either Ink4a-Arf or p53 in the muscles of limbs generated pleomorphic soft-tissue sarcomas.10 Subtype-specific diagnoses have improved dramati- cally for translocation-associated sarcomas (Table 1). Molecular methods, such as spectral karyotyping, fluo- rescent in situ hybridization, and real-time reverse tran- scription polymerase chain reaction for fusion tran- scripts are becoming more widely available to diagnostic laboratories. Mouse models have confirmed the causative relationship between the translocation- generated fusion protein and the sarcoma for three spe- cific types: myxoid liposarcoma, alveolar rhabdomyo- sarcoma, and synovial sarcoma.11-13 Other sarcomas also have discernible genetic back- grounds (Table 2). Patients with neurofibromatosis type I, from inherited mutation in the NF1 gene, are predisposed to the development of malignant periph- eral nerve sheath tumors. Patients with Ollier disease or Maffucci syndrome have multiple enchondromas with a high rate of malignant transformation to chondrosar- comas. Mouse models of Ollier disease, which use a va- riety of genetic derangements to effect increased Indian hedgehog signaling, have been used to study the pro- gression to chondrosarcoma.14 Patients with multiple osteochondromas, bearing germline mutations in EXT1 or EXT2, develop numerous metaphyseal osteochon- dromas and rarely a surface chondrosarcoma (1% to 3% lifetime risk per patient).15 The neoplastic character of two lesions, whose clon- ality has long been questioned, has recently been set- tled. Pigmented villonodular synovitis and aneurysmal bone cysts both share a unique pathophysiology char- acterized by a small amount (usually less than 10%) of Neither Dr. Jones nor any immediate family member has received anything of value from or owns stock in a com- mercial company or institution related directly or indi- rectly to the subject of this chapter. 193© 2011 American Academy of Orthopaedic Surgeons Orthopaedic Knowledge Update 10 2: S ys te m ic D is o rd er s Table 1 Sarcoma Translocations Sarcoma Chromosome Translocation Fusion Gene Alveolar rhabdomyosarcoma t(2;13)(q35;q14) PAX3-FKHR t(1;13)(q36;q14) PAX7-FKHR Alveolar soft-part sarcoma t(X;17)(p11;q25) TFE3-ASPL Aneurysmal bone cyst 17p3 rearrangement USP6 increase Clear cell sarcoma t(12;22)(q13;q12) EWS-ATF1 Congenital fibrosarcoma t(12;15)(p13;q25) ETV6-NTRK3 Dermatofibrosarcoma protuberans t(17;22)(q22;q13) COL1A1-PDGFB Desmoplastic small round cell tumor t(11;22)(p13;q11) EWS-WT1 Extraskeletal myxoid chondrosarcoma t(9;22)(q22;q12) EWS-CHN t(9;17)(q22;q11) TAF2N-CHN Ewing sarcoma family of tumors t(11;22)(q24;q12) EWS-FLI1 t(21;22)(q22;q12) EWS-ERG t(7;22)(p22;q12) EWS-ETV1 t(2;22)(q33;q12) EWS-E1AF t(17;22)(q12;q12) EWS-FEV Fibromyxoid sarcoma, low-grade t(7;16)(q33;p11) FUS-CREB3L2 t(11;16)(p11;p11) FUS-CREB3L1 Inflammatory myofibroblastic tumor t(1;2)(q22;p23) TPM3-ALK t(2;19)(p23;p13) TPM4-ALK Myxoid liposarcoma t(12;16)(q13;p11) FUS-DDIT3 t(12;22)(q13;q12) EWS-DDIT3 Pigmented villonodular synovitis 5q33 rearrangement CSF1 increase Synovial sarcoma t(X;18)(p11;q11) SYT-SSX1 SYT-SSX2 SYT-SSX4 Table 2 Sarcomas Associated With Genetic Predispositions to Cancer Heritable Syndrome Gene(s) Involved Associated Musculoskeletal Neoplasm Li Fraumeni p53 Osteosarcoma, pleiomorphic rhabdomyosarcoma, pleiomorphic undifferentiated sarcoma Congenital bilateral retinoblastoma RB1 Osteosarcoma Rothmund-Thomson RECQL4 Osteosarcoma Multiple hereditary exostoses EXT1, EXT2 Osteochondroma, secondary chondrosarcoma Neurofibromatosis (type I) NF1 Neurofibroma, malignant peripheral nerve sheath tumor McCune-Albright GNAS1 Fibrous dysplasia Ollier disease PTHR1 in minority Enchondromas Section 2: Systemic Disorders 194 Orthopaedic Knowledge Update 10 © 2011 American Academy of Orthopaedic Surgeons 2: System ic D iso rd ers cells in the tumor volume composed of neoplastic cells bearing specific chromosomal translocations: CSF-1 gene rearrangements16 and chromosome 17p3, USP6 gene rearrangements,17 respectively. The remaining cells in the lesion are not neoplastic, but are recruited to the neoplasm to create what is termed a landscape effect. Clinical Research Paradigms Surgically-related clinical research is focused on im- proving the quality and longevity of functional out- comes following limb-sparing resection of tumors. The previous standard outcomes instruments were the 1987 and 1993 Musculoskeletal Tumor Society outcome scores. Often, both scores are used in tandem. The for- mer score is joint specific and the latter score is more generalized. Both instruments use a physician-focused rather than a patient-focused approach. The Toronto Extremity Salvage Score is a patient- and function- focused outcome score that also is generalized and is not joint or limb specific. Most surgical studies in the literature related to sar- coma come from single centers or ad hoc collabora- tions between a few centers. There have been a few cross-Canadian and cross-European collaborative stud- ies, but more are needed. The study of sarcoma began with one of the first national collaborative registries, called the Bone Tumor Registry, which focused on mus- culoskeletal neoplasms. This registry was in operation from the 1920s through 1953, when data collection ceased. The current medicolegal environment and re- quirements of the Health Insurance Portability and Ac- countability Act (HIPAA) make it very unlikely that a similar contemporary registry will be established in the United States. There are collaborative groups that continue to study sarcoma, including the Children’s Oncology Group, Sarcoma Alliance for Research through Collab- oration, and the Radiation Therapy Oncology Group; however, these groups rarely conduct studies regarding surgical techniques or outcomes. Bone Lesions Patient Presentations Musculoskeletal neoplasms and lesions that mimic such neoplasms come to the attention of medical caregivers when a patient presents for treatment because of pain, a detected mass, a fracture, or when an imaging abnor- mality is noted during the evaluation for an unrelated disorder. This last group of incidentally noted lesions requires diligent management; however patience and serial imaging may confirm the latency of such lesions without the anxiety or expense created by investiga- tions using more complex modalities. Each of these four categoric presentations can overlap. Even inciden- tally noted lesions may be found to be symptomatic when the patient is probed with specific questions. These overlaps in the reason the patient seeks treatment can guide the development of a differential diagnosis. For example, a patient who presents with a fracture through a lesion who had experienced antecedent pain raises suspicion of different diagnoses than a patient whose fracture occurred through a previously asympto- matic lesion. Clinical Evaluation When taking the patient’s history, there should be a critical focus on discerning the patterns of typical or- thopaedic diagnoses. Specifically, the history of pain over time is crucial. Pain at rest or at night is indicative of biologic pain from the growth of a lesion in a bone. Pain with weight bearing or activity raises the suspicion of mechanical pain from poor structural integrity in the bone. Pain after activity is more indicative of an inflam- matory phenomenon than a neoplasm. The pace of dis- ease over time also can be informative. A mass that is present for years is unlikely to be an aggressive malig- nancy unless there has been a recent change in the pace of the disease process. A lytic lesion causing bone pain that evolves over hours or days is more likely infectious than neoplastic, whereas lesional pain evolving over weeks or months is more likely to indicate a neoplasm. The physical examination should be focused on rul- ing out alternate orthopaedic diagnoses. Identifying the precise location of the pain is critical. If pain does not colocalize with the mass or bone lesion identified with imaging, it should be determined if the pain fits typical referred pain locations, a radicular distribution, or a peripheral nerve distribution. Few laboratory tests provide useful diagnostic infor- mation. Important exceptions are inflammatory mark- ers in the setting of possible infections, lactate dehydro- genase in the setting of possible lymphoma of bone, serum and urine protein electrophoreses when there is concern about myeloma, alkaline phosphatase as a prognosticator for osteosarcoma, and specific tumor markers in the workup of metastatic carcinoma such as prostate-specific antigen. Imaging Plain radiography remains the diagnostic imaging mo- dality of choice for nearly all skeletal neoplasia. For le- sions located in areas that would be difficult to visual- ize with plain radiography, such as the sacrum and the scapulae, CT is the first alternative. These x-ray–based modalities demonstrate the matrix formed by the lesion and the zone of transition between the lesional tissue and host bone. Matrix types include bone (appearing as smooth mineralization), cartilage (appearing as stippled mineralization in rings and arcs), and fibro-osseous ma- trix (ground-glass appearance). The classic categorization of zones of transition be- tween lesion and host tissue was first described in 1980.18 Three such categories are in current use: latent lesions surrounded by a reactive cortical rim; active le- sions with an abrupt, easily discernible transition but no reactive rind (Figure 1); and aggressive lesions, with a broad, infiltrating border with the host. These classi- fications reflect the lesion’s presumed activity over Chapter 17: Musculoskeletal Oncology 195© 2011 American Academy of Orthopaedic Surgeons Orthopaedic Knowledge Update 10 2: S ys te m ic D is o rd er s time, but serial imaging remains the most definitive as- sessment of lesional behavior. There are few, if any, bone lesions that do not require at least a second set of imaging studies, separated in time by months, to con- firm latency. The location of the lesion in the bone also guides the differential diagnosis (Table 3). Most, but not all, le- sions have a predilection for the metaphyses near major growth centers of the skeleton. There are few differen- tial diagnoses for entirely epiphyseal lesions or those lo- cated in the small bones of the wrists and ankles. Sim- ilarly, few lesions will affect the diaphysis and spare the metaphysis. Staging is also performed by imaging, but requires distinct modalities. Local staging is achieved with MRI, which can best identify and localize any soft-tissue ex- tension of the lesion. Although some lesions, such as gi- ant cell tumor of bone, have characteristic appearances on MRI, this modality is primarily used for staging rather than diagnosis. For malignancies, systemic stag- Figure 1 A 35-year-old woman delayed seeking medical attention until she felt a pop and could not bear weight on her right knee. AP (A) and lateral (B) radiographs of the knee show a lytic lesion with a narrow zone of transition, but no reactive rind of cortical bone. Such a lesion-host interface is called an active border and is given a Lodwick A2 rat- ing. Lesions with active borders are usually in the category of benign aggressive bone lesions. C and D, MRI scans (T1-weighted coronal and T2-weighted axial, respectively) show a mass filling the proximal tibia, with subchondral fracture and tibial tubercle compromise. Incisional biopsy was consistent with giant cell tumor of bone. High-speed burr-enhanced intralesional excision was performed, followed by allograft reconstruction of the bone defect and reinforcement of the extensor mechanism. Section 2: Systemic Disorders 196 Orthopaedic Knowledge Update 10 © 2011 American Academy of Orthopaedic Surgeons 2: System ic D iso rd ers ing is required and usually includes a technetium Tc 99m total body bone scan and noncontrast CT of the chest to seek potential sites of metastasis. High suspi- cion must be maintained for false-negative bone scans in the setting of multiple myeloma or diffusely meta- static prostate carcinoma; the former for its lack of de- tection by bone scan, the latter because evenly in- creased uptake throughout whole sections of the skeleton can create a superscan effect, which can be av- eraged to appear negative. For myeloma, specifically, a skeletal survey is the preferred method for screening the skeleton. Benign bone tumors are often staged according to the Campanacci19 radiographic system, which was adapted from the Enneking clinical system. Each sys- tem includes grade 1 tumors, which are latent and sur- rounded by a reactive rind; grade 2 tumors, which are active but contained within at least a neocortex, if not the original cortex of the host bone; and grade 3 tu- mors, which include soft-tissue masses extending be- yond the cortex and not contained by the neocortex. Malignant bone neoplasms are often staged using the Enneking system, as adapted by the Musculoskeletal Tumor Society.20 Stage I are low-grade lesions, stage II are intermediate- or high-grade lesions, and stage III are lesions with demonstrable metastatic disease. For stage I and II lesions, an intracompartmental, A, or extra- compartmental, B, designation relating to the local ex- tent of the disease is applied. The more formal staging system of the American Joint Committee on Cancer (Table 4) is increasingly used as an alternative or ad- junct to the Enneking-Musculoskeletal Tumor Society staging system and is recommended for communication with oncologists and for central registry data entry. Iliac crest bone marrow biopsy also is included in the disease-specific staging systems for myeloma and the Ewing sarcoma family of tumors. Surgeons can fa- cilitate the use of this evaluation tool if the biopsy is performed with the patient under general anesthesia. Biopsy The purpose of biopsy is to obtain diagnostic tissue as well as specimens for tissue-banking and research. Di- agnostic tissue can be procured by fine-needle aspira- tion, core needle biopsy, incisional biopsy, or excisional biopsy and may be timed concurrent with the definitive surgery or long before it, depending on the clinical sce- nario. Biopsies are best performed by a team prepared to provide definitive treatment. Such interdisciplinary teams can best judge which lesions require biopsy and which biopsy method will be best suited to the patient’s potential diagnoses.21,22 Although few scenarios are safely managed with intraoperative frozen section diag- nosis followed by definitive management, obtaining Table 3 Bone Tumor Location Within the Bone Defines the Differential Diagnosis Epiphysis Metaphysis Diaphysis Chondroblastoma Clear cell chondrosarcoma Extension of giant cell tumor of bone Osteochondromas in Trevor disease Most common site for most bone neoplasms, primary or metastatic Fibrous dysplasia Ewing sarcoma Langerhans cell histiocytosis Osteoid osteoma Osteoblastoma Osteofibrous dysplasia/adamantinoma Lymphoma Metastatic carcinoma Myeloma Table 4 American Joint Committee on Cancer Staging System Stage Histologic Grade Size Location (Relative to Fascia) Systemic/Metastatic Disease Present IA Low < 5 cm Superficial or deep No IB Low ≥ 5 cm Superficial No IIA Low ≥ 5 cm Deep No IIB High < 5 cm Superficial or deep No IIC High ≥ 5 cm Superficial No III High ≥ 5 cm Deep No IV Any Any Any Yes Chapter 17: Musculoskeletal Oncology 197© 2011 American Academy of Orthopaedic Surgeons Orthopaedic Knowledge Update 10 2: S ys te m ic D is o rd er s frozen sections to confirm the adequacy of tissue is crit- ical to the performance of surgical incisional biopsies and thus requires a musculoskeletal pathologist. Biop- sies performed without considering definitive surgical options can have severe consequences caused by poor placement of the incision, violation of otherwise non- contaminated tissue compartments, or by spreading tu- mor cells by hematoma formation.21,22 Not all lesions should be biopsied. Asymptomatic, latent-appearing bone lesions that represent no signifi- cant risk of pathologic fracture based on their size and location should be monitored with serial imaging to confirm latency, rather than exposing patients to the risks of biopsy. Cartilaginous lesions should be biopsied only with the intent of confirming their cartilaginous character if aggressive treatments are indicated because grading of such lesions has been shown to be unsatis- factory, even among skilled pathologists.23 Hematoxylin and eosin staining is the pathologist’s primary diagnostic tool for bone neoplasms. Although immunohistochemical stains are used in specific scenar- ios, such as for small, round, blue cell-appearing le- sions, no specific diagnostic tests are available for most bone neoplasms. This situation places increased empha- sis on the experience of the interpreting pathologist. For small, round, blue cell tumors, several markers