Resection is required for all subtypes of osteosarcomas; however, the role of chemotherapy is dependent on the histologic grade of the tumor. Localized low-grade tumors, including low-grade central and parosteal osteosarcoma, are treated with surgery alone. Chemotherapy for periosteal osteosarcoma (intermediate-grade) is controversial, with most studies demonstrating no additional survival benefit with chemotherapy.[22,25] High-grade osteosarcoma, which includes conventional intramedullary, telangiectatic, small-cell, and high-grade surface osteosarcoma, is treated with a multidisciplinary approach consisting of chemotherapy and surgical resection. Before the adoption of chemotherapy, only 20% of patients achieved long-term survival. With the adoption of chemotherapy in the 1970s, the 5-year OS improved to approximately 60%; however, survival has improved minimally over the subsequent decades. Recent efforts have focused on understanding the influence of patient-specific and tumor-specific factors on oncologic outcomes in order to develop risk-stratified protocols to minimize treatment-related morbidity and maximize oncologic outcomes. International collaboration has been crucial to obtain statistically meaningful data with shorter study durations and multiple experimental arms. Finally, advances in implant design, surgical techniques and technology, and the integration of microvascular reconstructive surgeons as part of multidisciplinary surgical teams have helped optimize oncologic and functional outcomes.
Localized and Resectable Metastatic Disease
The standard approach for patients with localized or resectable metastatic high-grade osteosarcoma remains neoadjuvant chemotherapy, resection of all disease to achieve a complete surgical remission, and completion of adjuvant chemotherapy. Standard neoadjuvant chemotherapy consists of a three-drug regimen, with alternating cycles of doxorubicin, cisplatin, and high-dose methotrexate. Dose intensification and the addition of more agents to the standard protocol, tailored based on patient and treatment factors, have been the focus for many of the recent chemotherapy trials.
In an effort to improve outcomes, various agents were added to the standard therapy. Initially, ifosfamide was added to neoadjuvant and adjuvant regimens in a multiinstitutional trial; however, this demonstrated no benefit and was associated with higher toxicity. Subsequently, muramyl tripeptide phosphatidylethanolamine (MTP-PE), a monocyte and macrophage activator, was assessed in patients with localized and resectable high-grade, intramedullary osteosarcoma. The 6-year EFS was 66%, and the 6-year OS was 74%. However, the interpretation of these results, which demonstrated a statistically significant improvement in OS with MTP-PE, has been controversial because of the study design. Ultimately, MTP-PE was not approved by the Food and Drug Administration in the United States, but it is currently used in Europe.
The largest international trial in patients with osteosarcoma to date, EURAMOS-1 sought to evaluate the benefit of risk-stratified protocols for patients with high-grade osteosarcoma. The benefit of pegylated interferon-α-2b as maintenance therapy after the completion of adjuvant chemotherapy was investigated in patients with resectable localized or metastatic high-grade osteosarcoma of the extremity and axial skeleton who had a good histologic response (<10% viable tumor remaining). No difference was observed in EFS or OS between the patients who received maintenance pegylated interferon-α-2b. In an effort to improve the dismal survival associated with a poor histologic response (>10% viable tumor remaining) to neoadjuvant chemotherapy, ifosfamide and etoposide were added to the adjuvant chemotherapy regimen. Poor responders were randomized to receive methotrexate, doxorubicin, and cisplatin (MAP) or MAP plus ifosfamide and etoposide (MAPIE) chemotherapy in the adjuvant setting. No difference was observed in a 3-year EFS (55% for MAP and 53% for MAPIE) or a 3-year OS (72% MAP and 77% MAPIE) between the groups. The MAPIE regimen was associated with increased toxicity, decreased total received doses, and higher rates of secondary malignancies, thus not supporting the addition of ifosfamide and etoposide to adjuvant regimens in poor responders. In response to stagnation with dose intensification protocols, future trials adding targeted chemotherapeutics and immunotherapies with robust preclinical evidence to the MAP backbone are currently being explored.
Metastatic and Relapsed Osteosarcoma
Between 14% and 23% of patients with high-grade osteosarcoma will present with radiographically detectable metastatic disease,[16,30,31] which is associated with a 5-year EFS of 28% and a 5-year OS of 45%. Lung metastases comprise the overwhelming majority of the metastatic burden in patients with osteosarcoma (85% of metastatic lesions), whereas skeletal metastases account for 21% and other sites only 9%. Risk factors for metastatic disease at presentation include age >60, axial location, large size, lowest quartile of income, and the lowest composite socioeconomic status score. The number and location of pulmonary metastases have an effect on survival with more numerous and centrally located metastases associated with worse outcomes.
Seventy-five percent of patients with metastatic disease at presentation will have pulmonary, oligometastatic disease, of which 50% of these metastases are resectable. Treatment of oligometastatic disease includes neoadjuvant chemotherapy, primary tumor resection, adjuvant chemotherapy, and then pulmonary metastasectomy. Resection of all diseases with negative margins is required for long-term survival in patients with isolated pulmonary metastases, an approach that achieves long-term survival of up to 40%. By contrast, patients with pulmonary and extrapulmonary metastatic disease have a worse prognosis.
Forty-five percent of patients with initially localized osteosarcoma will relapse, of which 81% of relapses are isolated to the lungs. Prognosis in patients with relapsed pulmonary metastases is poorer when associated with shorter disease-free intervals,[31,35] central and/or bilateral pulmonary metastases,[33–35] and more numerous pulmonary metastases.[7,33] Among these patients, 63% are candidates for pulmonary metastasectomy which is associated with a 29% 5-year OS. Metastasectomy can be repeated as necessary.
LR most commonly occurs in the soft tissues and within the first 2 years after surgery. LR is associated with inadequate surgical margins and/or a poor histologic response. When LR is associated with a distant recurrence, OS is dismal; however, a negative margin resection of isolated locally recurrent disease is associated with improved long-term survival (up to 25% of a 5-year OS) and is generally recommended. Given the poor prognosis in patients with extensive or unresectable recurrences, radiation and second-line chemotherapy are viable salvage options.
Unresectable metastatic disease is managed primarily with chemotherapy, whereas surgery and radiation serve palliative roles. Progressive metastatic disease is typically treated on clinical trials. Novel targeted therapies, in particular tyrosine kinase inhibitors, have gained attention recently. Specifically, 30% of patients had progression-free survival ≥6 months with sorafenib, and 44% of patients had a 4-month progression-free survival with regorafenib, although no complete responses were seen with either drug.
Currently, radiation therapy is used only in the palliative setting. Symptomatic relief may be obtained with site-specific radiation for unresectable LRs or symptomatic metastatic lesions. In addition, whole lung irradiation with chemotherapy may provide temporary benefit in patients with pulmonary metastases. Finally, newer techniques, including stereotactic body radiation therapy and radiofrequency ablation, may have evolving roles in the local control of metastatic lesions.
Local control through negative margin surgical resection is crucial for long-term survival. Advances in preoperative imaging and reconstructive options have made durable and functional limb salvage possible with acceptable LR rates (5% to 10%) in most patients.[8,36,39] Intralesional and positive margin resections are associated with higher rates of LR and poorer survival. However, LR may occur even after wide surgical margins suggesting that the propensity for LR may be inherent to tumor biology and not margin status.[39,40] Although at least a 2-cm osseous margin is frequently suggested, no consensus regarding the definition of "adequate surgical margins" exists. More recent evidence suggests that this "wide" bony margin may not be critical, and by contrast, a "negative" margin with a cuff of normal tissue may be all that is required.
Less common amputation may be necessary when dictated by tumor or patient factors. Although limb salvage is associated with slightly superior physical function compared with amputation, both groups have similar psychosocial outcomes with high levels of quality of life, self-esteem, social support, social participation, and economic independence. Amputation and limb salvage both have notable and unique disadvantages. Amputees may experience phantom limb pain, prosthetic issues, residual limb complications, and a lifelong need for assistive devices. In comparison, limb salvage patients may have lifelong activity restrictions and are at risk for prosthetic complications, in particular, mechanical failure and infection. These complications may necessitate secondary amputation in approximately 5% to 10% of patients. Given similar outcomes, the decision to proceed with limb salvage or amputation should take into consideration the patient's lifestyle and activity goals.
The periarticular predilection of osteosarcoma often necessitates joint reconstruction, the options for which include endoprostheses, allograft-prosthetic composites, osteoarticular allografts, and arthrodesis. Endoprosthetic reconstructions provide several advantages, including implant modularity and early weight-bearing, but have notable complications. Henderson et al classified the major modes of failure for endoprostheses, which include infection (34.1%), aseptic loosening (19.1%), structural failure and tumor progression (both 17.4%), and soft-tissue failure (12.0%). Certain modes of failure are associated with unique anatomic locations. Highly constrained hinged joints (knee and elbow) are at risk for aseptic loosening, whereas instability is a challenge with polyaxial joints (hip and shoulder). Allograft-prosthetic composites allow for restoration of bone stock and soft-tissue reattachments, but are subject to the unique complications of allografts. Finally, osteoarticular allografts remain an option; however, they are associated with high rates of mechanical failure that limit their clinical use.
In skeletally immature patients, resection of the physis may lead to a notable limb-length discrepancy, which presents a unique reconstructive challenge. In response, surgeons developed expandable endoprostheses. With recent advances, noninvasive, in-office expansion is possible (Figure 4). The long-term results of 124 expandable endoprostheses were reviewed and demonstrated high rates of limb salvage, good functional outcomes, and excellent maintenance of limb length. However, complications were frequent, in particular, subluxation of the hip in 77% of patients with proximal femoral replacements, aseptic loosening in 52% of distal femoral and proximal tibial replacements, and infection in 34% of proximal tibia replacements.
Radiographs showing noninvasive expandable endoprostheses. To compensate for loss of the physis in skeletally immature children with notable growth remaining, a noninvasive growing prosthesis may be used. The telescoping magnetic-drive growing mechanism allows lengthening to be done in the clinic. Radiographs (A) early in the lengthening process and (B) at skeletal maturity demonstrate symmetric leg lengths with notable lengthening through the diaphyseal segment of the implant.
Isolated diaphyseal involvement, more common with periosteal osteosarcoma, may be treated with an intercalary resection. Allograft reconstruction is the most common technique for the management of intercalary resection defects; however, it is associated with high rates of complications requiring secondary procedures, including allograft-host junction nonunion, fracture, and infection (40%, 29%, and 14%, respectively). Augmentation of intercalary allografts with vascularized fibula grafts is an effective strategy to minimize these complications, and in the upper extremity and pediatric patients, vascularized fibula grafts may be used alone.
When dictated by tumor characteristics or patient preference, ablative procedures may be necessary to achieve oncologic goals at the index procedure or in the setting of LR or failed limb salvage. Rotationplasty, even in skeletally mature patients, may be an option to salvage knee function in a patient who would otherwise require an above knee amputation. Rotationplasty is associated with good functional results, but it is not without risks, including vascular compromise. Finally, novel surgical techniques in amputee care, including targeted muscle reinnervation and osseointegrated percutaneous prostheses, show promise for improving functional outcomes and postamputation pain.
Although previously controversial, patients presenting with or sustaining an in-treatment pathologic fracture through an osteosarcoma can be successfully managed with definitive limb salvage (when appropriate) rather than an amputation. Nonsurgical treatment with a sling, cast, or a temporary spanning external fixator outside the field of surgery is the treatment of choice when possible, although the recommended treatment is dependent on the involved bone and fracture characteristics. Patients are treated with neoadjuvant chemotherapy followed by limb salvage when feasible. Pathologic fractures are associated with higher rates of amputation and extraarticular resections. However, among patients undergoing limb salvage, Ferguson et al demonstrated no difference in LR rates compared with patients without a pathologic fracture. Although there was no difference in LR, patients who sustained a pathologic fracture had a worse 5-year OS (41% vs 60%). This supports the theory that a pathologic fracture is an expression of aggressive tumor biology, and therefore, decreased survival is independent of the surgical treatment and instead innate to the disease.
Finally, image-based surgical technology has been adapted for oncologic purposes to improve surgeon confidence in resection margins and improve the reliability of complex reconstructive techniques. In particular, patient-specific cutting guides and computer navigation have been shown to improve the accuracy of complex osteotomies which allows 3D-printed implants to be manufactured to precisely fit the resultant defect (Figure 15, Supplemental Digital Content, https://links.lww.com/JAAOS/A690). Although the availability and use of these techniques has yet to become widespread, there is a growing body of evidence that surgical adjuncts improve accuracy, with the hope that this translates to improved oncologic and functional results.
J Am Acad Orthop Surg. 2021;29(20):e993-e1004. © 2021 American Academy of Orthopaedic Surgeons