Limb preservation surgery has gained acceptance as a viable alternative to amputation for the treatment of extremity bone tumors in the growing child. There are several options for reconstructing the potential loss of a physis and the defect created by tumor excision. Metallic endoprosthesis, massive allograft, and allograft-prosthesis composites have been described in the skeletally immature population. With the development of expandable prostheses, even those far from skeletal maturity may be candidates for limb salvage. However, improvements in the literature are needed, including reporting surgical and functional outcomes in a rigorous manner, specific to age, anatomic location, and reconstruction.
Key points
- •
Limb salvage may be a viable alternative to amputation in even the youngest of pediatric patients with an extremity bone tumor.
- •
Expandable metallic endoprosthesis can achieve meaningful growth, resulting in little to no limb length discrepancy for skeletally immature patients by the end of their growth.
- •
Many case series have been presented in the literature, but there is no evidence graded level II or higher, limiting the ability to identify 1 construct as superior to another.
- •
Failure rates in the pediatric population of the 3 main forms of limb salvage constructs (metallic endoprosthesis, allograft, allograft-prosthetic composite) are high, but comparable with failure rates reported in adults.
- •
Function after limb salvage for bone tumor in the growing child has been inadequately reported in the literature.
Historical background
The prognosis of bone sarcoma of the extremities in the pediatric patient was historically described as poor, hopeless, or grave. In the first half of the twentieth century, the only curative procedure was amputation, which provided a reported 5-year survival of 12% to 23%. Amputation was not available for all patients; rather, it was reserved for patients who met appropriate criteria, which included no evidence of metastatic disease. With such a bleak prognosis, the idea of limb salvage seemed implausible.
In 1975, several studies were published that changed the landscape in the treatment of extremity sarcoma. Chemotherapy and radiation began to have an impact on the course of the disease, improving 5-year survival outcomes from 20% to 65%. As late as 1979, Copeland and Sutow noted that “following diagnostic biopsy, amputation of the extremities remains the treatment of choice. In selected cases, limb-saving radical en bloc resection may surface.” However, by 1986, Simon and colleagues reported on their results in 227 patients younger than 30 years with osteosarcoma; they reported no survival difference in 5-year follow-up between those patients who underwent amputation versus limb salvage. With this improvement in survival afforded by chemotherapy and radiation, the treatment paradigm shifted from radical amputation to en bloc resection and reconstruction of the resultant defect. Limb salvage is the prevailing treatment of most sarcomas of the extremities, with 90% to 95% of skeletally mature patients with primary bone tumors receiving some fashion of limb salvage surgery.
Although the overall trend in orthopedic oncology moved from amputation to limb salvage, the treatment of the pediatric patient with primary bone tumor remained controversial. This is partly because of the generally high complication rates after tumor resection and subsequent reconstruction and partly because of the unique issues related to pediatric patients when compared with their adult counterparts. With the rate of failure requiring operative revision nearing 25% in adults, some orthopedic oncologists have questioned the appropriateness of subjecting children with bone sarcoma to a lifetime of surgical complications and revisions. In addition, surgical resection of tumor in the skeletally immature patient is often complicated by the loss of a physis, with a resultant potential for clinically relevant leg length discrepancy. For these reasons, some tumor surgeons maintained that children with a malignant bone tumor of the lower extremity, especially those far from skeletal maturity, might be best served by a single amputation surgery with subsequent prosthesis fitting. From this point of view, amputation would provide the most consistent outcome with the least amount of surgeries and complications. In a review article published in Journal of the American Academy of Orthopedic Surgery in 2003, DiCaprio and Freidlaender list “immature skeletal age with a predicted leg length discrepancy greater than 8 cm” as a relative contraindication to limb salvage.
However, an increasing body of literature has reported limb salvage techniques for the skeletally immature, including patients younger than 5 years. Expandable prostheses have been developed and modified to address the issue of leg length discrepancy, and both surgical and functional outcome studies have been performed that suggest that limb salvage might be a preferable alternative to amputation in the skeletally immature patient with a primary bone tumor.
Historical background
The prognosis of bone sarcoma of the extremities in the pediatric patient was historically described as poor, hopeless, or grave. In the first half of the twentieth century, the only curative procedure was amputation, which provided a reported 5-year survival of 12% to 23%. Amputation was not available for all patients; rather, it was reserved for patients who met appropriate criteria, which included no evidence of metastatic disease. With such a bleak prognosis, the idea of limb salvage seemed implausible.
In 1975, several studies were published that changed the landscape in the treatment of extremity sarcoma. Chemotherapy and radiation began to have an impact on the course of the disease, improving 5-year survival outcomes from 20% to 65%. As late as 1979, Copeland and Sutow noted that “following diagnostic biopsy, amputation of the extremities remains the treatment of choice. In selected cases, limb-saving radical en bloc resection may surface.” However, by 1986, Simon and colleagues reported on their results in 227 patients younger than 30 years with osteosarcoma; they reported no survival difference in 5-year follow-up between those patients who underwent amputation versus limb salvage. With this improvement in survival afforded by chemotherapy and radiation, the treatment paradigm shifted from radical amputation to en bloc resection and reconstruction of the resultant defect. Limb salvage is the prevailing treatment of most sarcomas of the extremities, with 90% to 95% of skeletally mature patients with primary bone tumors receiving some fashion of limb salvage surgery.
Although the overall trend in orthopedic oncology moved from amputation to limb salvage, the treatment of the pediatric patient with primary bone tumor remained controversial. This is partly because of the generally high complication rates after tumor resection and subsequent reconstruction and partly because of the unique issues related to pediatric patients when compared with their adult counterparts. With the rate of failure requiring operative revision nearing 25% in adults, some orthopedic oncologists have questioned the appropriateness of subjecting children with bone sarcoma to a lifetime of surgical complications and revisions. In addition, surgical resection of tumor in the skeletally immature patient is often complicated by the loss of a physis, with a resultant potential for clinically relevant leg length discrepancy. For these reasons, some tumor surgeons maintained that children with a malignant bone tumor of the lower extremity, especially those far from skeletal maturity, might be best served by a single amputation surgery with subsequent prosthesis fitting. From this point of view, amputation would provide the most consistent outcome with the least amount of surgeries and complications. In a review article published in Journal of the American Academy of Orthopedic Surgery in 2003, DiCaprio and Freidlaender list “immature skeletal age with a predicted leg length discrepancy greater than 8 cm” as a relative contraindication to limb salvage.
However, an increasing body of literature has reported limb salvage techniques for the skeletally immature, including patients younger than 5 years. Expandable prostheses have been developed and modified to address the issue of leg length discrepancy, and both surgical and functional outcome studies have been performed that suggest that limb salvage might be a preferable alternative to amputation in the skeletally immature patient with a primary bone tumor.
State of the literature
There are several features inherent to pediatric orthopedic oncology that hamper the development of a robust literature. First, the relative rarity of pediatric bone tumor limits the amount of patients any single treatment center may see over the course of a reasonable timespan. When segregated for anatomic location of the tumor, the number of patients can become even more limited. For example, in a 10-year span, many large regional cancer centers may have only a few pediatric patients with a distal femur tumor that was treated with en bloc resection and a specific limb salvage option, such as endoprosthesis, allograft, or allograft-prosthetic composite (APC). By the time a sufficient number of patients have been accumulated, the interventions may have been modified and no longer be comparable. Second, tumor resections are not homogenous. Outcomes also depend on soft tissue maintenance, nerve function, and health of the patient. However, these factors cannot be randomized or controlled, because they are subject to the specifics of the tumor and the patient; strict exclusion criteria for study would only lead to further restriction on patient number, often to the point of reporting on patients on a case-by-case basis.
In a recent review of the literature, an attempt was made to investigate the surgical and functional outcomes in pediatric patients after en bloc bone sarcoma excision followed by limb salvage reconstruction with an endoprosthesis, an allograft, or an allograft-prosthesis reconstruction. The study segregated each limb salvage type by anatomic location, reporting on the proximal femur, total femur, distal femur, and the proximal tibia individually. The state of the literature was noted to be limited. No study rose above level IV evidence. In total, 62 articles were found that clearly reported on the surgical and functional outcomes of pediatric patients after limb salvage surgery for bone tumor at a specific anatomic location. No study had more than 40 patients, and 82% of the studies had 10 or fewer patients. Twenty-six of the 62 studies (42%), reported on 1 subject per described anatomic region. One hundred and seventy-five papers were excluded because of overinclusion and inability to reliably parse the data. That is, most studies combined pediatric patients with adult and elderly patients or combined hip, knee, and shoulder patients together as 1 group. Future research studies are needed that segregate pediatric patients from adults; more specifically, they should distinguish between those far from skeletal maturity versus patients in their peak of growth versus those patients nearing physeal closure. In addition, anatomically specific reporting should become the standard, because each region has unique challenges and surgical considerations.
Current alternatives to limb salvage
Amputation
Even in an era of expanding indications for limb salvage, amputation as the primary surgical modality in the treatment of pediatric extremity bone tumor remains a common intervention. Schrager and colleagues analyzed the registry data from the National Cancer Institute’s SEER (Surveillance Epidemiology and End Results) Program from 1988 to 2007. These investigators compared the incidence of amputation versus limb salvage in 890 patients with osteosarcoma younger than 20 years. They found that 66.3% of the reported sample underwent limb salvage, whereas 33.7% had amputation. A study using the same database noted that patients with an age of 5 years or less were more likely to undergo amputation than their 6-year-old to 19-year-old counterparts, with the former receiving amputation in 55% of cases versus 27% for the latter. This finding suggests that young age continues to be an exclusion criterion for many surgeons when considering limb salvage. This situation may be because of the clinically significant longitudinal and cross-sectional growth remaining in the child and the understanding that limb salvage means multiple surgeries throughout adolescence, even if a growing prosthesis was successfully used.
In addition to young age, traditional indications for amputation include patients with tumors whose resections would necessitate sacrifice of a major neurovascular bundle; tumors of the distal tibia; and patients in areas of the world that lack the economic resources to provide a complex implant. Although these indications are being challenged by an increasing body of literature showing alternatives to amputation, amputation remains a major tool for many orthopedic oncologists, as shown by the percentages noted earlier.
Rotationplasty
Because most of the primary bone tumors occur about the hip and knee, amputation often requires an above-knee amputation or hip disarticulation. This option imparts a significant increase in oxygen demand while walking when compared with a below-knee amputation and has been linked to less favorable psychosocial outcomes. For cases in which an above-knee amputation would be required to achieve tumor resection, rotationplasty is available as a means to effectively convert an above-knee amputation to a below-knee amputation.
First described by Borggreve in the 1930s and expanded by Van Nes in the 1950s to treat proximal focal femoral deficiency, rotationplasty involves the resection of the distal femur and proximal tibia, with subsequent tibiofemoral osteosynthesis, after rotating the distal tibia and foot 180°. This technique allows the residual distal tibia to augment the length of the foreshortened femur and uses the reversed but functional ankle to operate as a knee. The prosthesis fits on the foot and the patient functions as a below-knee amputee.
When considered for the growing child, rotationplasty offers a form of a growing reconstruction. The preserved proximal femoral physis remains and the distal tibial physis is preserved as well, providing potential growth at both ends of the newly fashioned femorotibial composite. Although the transferred distal tibial physis does not have the same growth potential as the contralateral distal femoral physis, the growth potential that is preserved allows procedures such as contralateral epiphysiodesis to be reasonably considered. The literature on rotationplasty is understandably limited to small case series, but reports of interventions in patients as young as 14 months old show the potential of this surgical alternative.
Surgical complications include the standard orthopedic oncology issues, such as infection, local recurrence, and pathologic fracture. Rotationplasty also may be subject to malrotation as well as failure of the vascular anastomosis. In a series of 25 pediatric patients receiving rotationplasty for tumor, Sawamura and colleagues reported 3 (12%) vascular complications requiring subsequent amputation. Six additional patients (24%) required subsequent surgical intervention for complications, including sciatic nerve palsy because of hematoma, distal tibia fracture, nonunion of the osteosynthesis, slipped capital femoral epiphysis, and wound complications.
When considering rotationplasty for the pediatric patient with a complex sarcoma about the knee, parents and surgeons must consider more than just the reconstruction and surgical complications. Amputation is a disfigurement in itself, and rotationplasty potentially magnifies this effect by altering the morphology of the extremity, beyond mere ablation. Forni and colleagues studied the emotional acceptance of children who underwent this procedure. These investigators found relational and emotion difficulty in adolescence after rotationplasty, which abated as the patient grew into adulthood. On the other hand, in a case series comparing rotationplasty with endoprosthetic reconstruction for treatment of tumors about the knee, Hillmann and colleagues reported equivalent functional scores between the groups, as measured by the Musculoskeletal Tumor Society (MSTS) questionnaire; moreover, significantly fewer of the rotationplasty patients required assistive walking devices, fewer had limiting pain, and more rotationplasty patients resumed hobbies and sports than their endoprosthesis-treated counterparts.
Because of the limitations of the literature, secondary to the rare nature of sarcoma and the even more limited number of rotationplasties, the outcomes of rotationplasty remain incompletely described. However, given the moderate support in the literature, rotationplasty remains a viable option in select cases, pending a thorough and frank discussion between the patient’s family and the surgeon.
Limb salvage: general principles for the skeletally immature
Initial efforts in limb salvage surgery for bone tumors were sporadic, controversial, and varied. Early attempts that were described included endoprostheses made of vitallium, ceramic, polyethylene, and cobalt-chrome. Alternatively, cadaveric allograft and autogenous cortical bone grafting were used. Some surgeons spanned the affected joint with a fusion construct, hoping to maintain length of the limb to preserve distal function, whereas more ambitious surgeons sought to restore joint function in addition to saving the limb. However, application of these preliminary techniques to the pediatric population was largely avoided. In 1976, Scales broke with this tradition, describing an expandable prosthesis designed specifically for the pediatric patient. Developed at the Division of Biomedical Engineering in Stanmore, England, the extendible endoprosthesis sought to reconstruct the posttumor resection defect and also to address the issue of anticipated limb length discrepancy after physis excision. Over the ensuing 40 years, the literature has come to reflect the increasing acceptance of limb salvage after bone tumor resection in the pediatric population, although considerable variation in reconstruction design persists.
Selecting a Prosthesis
Age-related concerns for the lower extremity
When considering the options for reconstruction after bone tumor resection in the lower extremity of a growing child, 3 distinct subpopulations based on age and skeletal maturity must be appreciated. The first subpopulation contains those patients nearing skeletal maturity, typically in the age ranges of 14 to 17 years. If the remaining anticipated growth of the resected physis is 2 cm or less, an adult implant may be used. The limb length discrepancy of this amount may be well tolerated without need to augment the limb with an external shoe lift. Alternatively, the adult implant can be selected such that the operated limb is lengthened intraoperatively to the expected adult length, allowing the contralateral side to catch up over time. The amount of overcorrection tolerated depends on the starting size of the limb and is limited by the potential to induce sciatic neuropraxia or loss of motion at the knee, secondary to the acute length change.
The second group of patients based on age is the skeletally immature with roughly 2 to 6 cm of growth remaining at the resected physis. This intermediate group may be addressed from various perspectives. Although the anticipated leg length discrepancy would be too great to simply overcorrect the affected side with a long implant, a combination of ipsilateral overcorrection with contralateral epiphysiodesis may result in acceptable limb length equality. The benefit of this approach lies in the fact that an adult implant is placed during the index procedure, forgoing the need to transition from a growing prosthesis to an adult implant once skeletal maturity is achieved. Alternatively, a growing prosthesis may be used. This option has several potential advantages over initial placement of an adult implant. An expandable prosthesis, if it functions properly, can be used to match the growth of the contralateral side. Although growth prediction provides a reasonable expectation of ultimate limb length, it is only an estimate and prone to error, especially in the face of chemotherapy and local radiation, which may affect physeal growth. Also, a growing prosthesis permits the greatest potential final patient height, as a result of preclusion of contralateral epiphysiodesis, which may be a concern in some cases.
The third and final subpopulation to consider in regards to age is the group far from skeletal maturity. Ranging from 2 to 8 years old, the only reliable option these patients have for limb salvage is a growing endoprosthesis. Showing the potential of limb salvage, reports have been published describing a proximal femur replacement in a 3-year-old, a total femur in a 2-year-old, distal femur replacement in a 5-year-old, and a proximal tibia replacement in a 4-year-old. At this extreme of young age, not only must longitudinal growth be considered but cross-sectional growth must also be taken into account. For the patient younger than 8 years to have successful limb salvage surgery, taking them to skeletal maturity with limb length equality and a functional limb, more than 1 growing prosthesis is likely required. With additional surgery for final conversion to an adult implant, limb salvage in this subpopulation requires a major commitment among the patient, the family, and the surgeon.
Age-related concerns for the upper extremity
Tolerance for limb length inequality is a major concern in the lower extremity, as a consequence of bipedal human ambulation, but limb length inequality in the upper extremity is not without its considerations. Moderate amount of limb length inequality is well tolerated in the upper extremity, and the patient can easily accept several centimeters of difference. This situation allows the surgeon to choose a final prosthesis based on the structural needs of the patient, whether that entails a metallic endoprosthesis (ME), allograft, or APC.
However, when performing a resection in the very young, a growing prosthesis in the humerus may be appropriate, to provide an upper limb that is morphologically similar to the contralateral side. However, these reconstructions are rare, because of the limited occurrence of sarcoma in general, and the even more restricted occurrence in the very young and in the upper extremity. Case reports have been published, but no substantial case series has been presented.
Metal Endoprostheses: Historical Background
Growing prosthesis
By 1992, Unwin and Walker were able to report on the outcomes of 168 skeletally immature patients who had bone tumors treated with en bloc resection and reconstruction with an extendible Stanmore custom ME. During the period reported, from 1976 to 1992, these investigators noted 4 distinct designs in the growing mechanism. The initial design used a worm-drive screw to extend the telescoping titanium components. That design was switched to a ball-bearing mechanism in 1982, because of reported mechanical failure of the device, attributed to difficulties associated with the manufacturing process. The ball-bearing mechanism was replaced in 1988 by a C-collar mechanism, in which expansion required open dissection to the area of the telescoped components, manual longitudinal traction, and the subsequent interposition of the C-collar spacer to maintain the gained length between the 2 components. This design was then replaced in 1992 with a minimally invasive device that used a mechanism similar to the original worm-screw design. An Allen wrench could be percutaneously inserted to turn the drive that engaged the screw mechanism. This history reflects the overall trend in expandable endoprosthetic design and use. The original attempts were limited by manufacturing capabilities and required invasive, open surgical procedures for every lengthening. As manufacturing techniques improved, the design could return to the screw mechanism and minimally invasive techniques could be used.
In 2006, Gupta and colleagues reported on the next step of development from the Stanmore group: the noninvasive growing prosthesis (Stanmore, Elstree, United Kingdom). This latest generation endoprosthesis, first used in 2002, houses a rare-earth magnetic disk within the endoprosthesis, which connects to a gearbox, which is in turn connected to a power screw. Expansion of the prosthesis is achieved when the extremity is placed in the core of a circular external drive unit. This external drive unit induces a rotating magnetic field that acts on the magnetic disk mechanism noninvasively within the endoprosthesis. The rotating magnetic field pulls the magnetic disks within the prosthesis, engaging the gearbox, which rotates the power screw, leading to the predictable lengthening of the prosthesis. The amount of potential expansion is determined by the size of the telescoping mechanism. Three lengths are available for the proximal femur, distal femur, and proximal tibia endoprosthesis, which provide 50 mm, 70 mm, or 90 mm of available growth, determined by the remaining growth of the child and the length of resection. The advantages of this latest generation of expandable prosthesis include reduced risk of infection because of noninvasive expansions, outpatient expansion without need for anesthesia, and precise and reversible lengthening. The specific disadvantages relative to other endoprostheses include high cost and limited availability in nonspecialized centers. Figs. 1–4 show a distal femur implant, the magnetic expansion device, and radiographic progression of expansion in an implanted total femur Stanmore device.
Alternatives to the Stanmore design have been developed and used. The Repiphysis expandable prosthesis is an example ( Figs. 5 and 6 ). Originally marketed as the Phenix prosthesis, the Repiphysis is a noninvasively expandable titanium endoprosthesis (Wright Medical, Arlington, TX). The expansion mechanism is housed within the tubular portion of the prosthesis and works by releasing stored energy from compressed springs within the mechanism. The spring is held compressed by 2 tubes, made of polyether ethylketone. When expansion is desired, the tubes are heated under the guidance of fluoroscopy by induction from an external magnetic coil; the heat melts the polymer and the coiled spring is released, creating the force to expand the tubular insert. However, general acceptance of the Repiphysis system has been limited because of concerns regarding the high rate of complications. Although limb salvage in general is plagued with failure rates in the 25% to 33% range, the Repiphysis system has been noted to have failure rates in excess of 50%. The group from Rush University reported, over the course of 2 studies, a failure rate of 58% (15 of 26 patients), with associated bone loss around the prosthesis often limiting revision options. Likewise, Staals and colleagues reported that 9 of the 10 patients they treated with Repiphysis reconstruction at the distal femur required revision surgery: 8 for implant breakage and 1 for aseptic loosening. In addition, these investigators noted that expansion lengthening was erratic and unpredictable. They cautioned against use of the device.
As an example of the innovative nature of endoprosthesis design in response to documented complications, Webber and Seidel combined the Compress compliant prestress device (Biomet, Warsaw, IN) with the Repiphysis endoprosthesis. Although these investigators presented only a single case report, and this has not been adopted or documented in any substantial number, this modification shows the ongoing adaptation occurring in pediatric limb salvage surgery. This will unlikely solve the high rate of complications associated with the Repiphysis, but the lesson in understanding and addressing the modes of failure of a device is worthy of note.
The MUTARS (Modular Universal Tumour and Revision System) shows additional innovation in expansion technology. In their Xpand prosthesis line, growth may be achieved through both mechanical as well as biological means. The mechanical device uses a high-frequency transmitter to activate the internal actuator within the prosthesis; to achieve biological growth, the stem of the prosthesis can be replaced with a growing intramedullary rod system that lengthens the host bone through distraction osteogenesis.
Although not exhaustive by any means, the synopsis given earlier reflects the early trials and subsequent innovations made in expandable prosthesis design. There are certainly other prostheses that should be considered in a complete history, such as the Lewis expandable adjustable prosthesis, which was the first widely used expandable device used in the United States. In addition to the noninvasive models noted earlier, minimally invasive expansion designs by Stryker and Biomet are available for use ( Figs. 7 and 8 ). A list of the current expandable MEs is presented in Table 1 .
