|Home | About | Journals | Submit | Contact Us | Français|
Surgical treatment and reconstruction of a pediatric patient with a bone malignancy should consider many patient and tumor specific factors. Surgical treatment should be geared first and foremost towards obtaining wide margins. To that end the options can include amputation, rotationplasty and prosthetic reconstruction. Advances in adjuvant chemotherapy for musculoskeletal malignancy in pediatric patients has increased acceptance of limb-salvage procedures as a viable option for treatment, whereas limb ablation was formerly the only acceptable means for attaining disease eradication. The advent of the expandable prosthesis has gained significant interest due to the appeal of improved cosmesis and potential for equal limb length at skeletal maturity. The latest generation implants allow for non-invasive lengthening with an outpatient procedure and are generally very well-tolerated by the patient. Review of current literature demonstrates that this procedure has generally good patient reported outcomes but has a high complication rate. Aseptic loosening and mechanical dysfunction are common modes of failure and often necessitate one or more large revision surgeries. Further improvement in implant design and biomaterials may decrease the incidence of these complications and promising work in these areas is ongoing. When discussing this specific option, patients and family should be counseled regarding the likelihood of future surgeries to manage the expected complications.
Primary bone malignancy, although an uncommon form of cancer, is most often encountered in the pediatric population.3,42The most common forms encountered are osteosarcoma and Ewing's sarcoma.26,28,30 Adequate treatment of these malignancies relies on a multidisci-plinary approach; with the surgeon's goal to obtain a wide resection of the tumor.40 Advances in diagnostics and chemotherapeutic treatment regimens have led to improved survival.4,40 In addition, they offer improved local control which has made limb salvage surgery an option for most patients.39,40
These bone tumors are frequently found at the physes of long bones, with the knee being the most common location.3,28,30,42 Therefore, the surgeon must consider that any surgical resection, especially limb salvage techniques, will cause a limb length discrepancy from physeal resection.10 This discrepancy must somehow be reconciled with surgical techniques to approximate equal leg lengths at skeletal maturity. This historically has been done with a combination of acute lengthening and contralateral epiphysiodesis. Currently there are several manufacturers who now offer options for prosthetic reconstruction with an expandable implant. This option offers the patient good function with hope for an equal leg length at maturity.10 In this paper we will review the indications for expanding endoprosthesis and alternative modes of treatment. We also will review the principles of surgical technique and available outcome data pertinent to the expanding endoprosthesis in the treatment of pediatric musculoskeletal malignancy.
When evaluating the pediatric patient with a primary bone malignancy the primary goal of the treating surgeon is wide resection of the lesion.40 This may be accomplished with several surgical options: limb ablation, rotationplasty and reconstruction.9,18-20,40
Limb ablation, amputation or disarticulation, was once considered the only acceptable method for eradication of malignancy in the lower extremity. Multimodal therapy, especially improved chemotherapeutic regimens, has largely made limb salvage techniques the standard of care for the large majority of tumors.4,34,39 A recent study comparing surgical techniques in patients with osteosarcoma and Ewing's sarcoma has noted improved Musculoskeletal Tumor Society (MSTS) scores in patients undergoing limb salvage compared with amputation.21 However, depending on specific patient and tumor circumstances, limb ablation remains a good option. Simon et al. concluded in their analysis of 22 patients that those treated with amputation were least worried about damaging their limb, but had more difficulties with navigating uneven terrain.39 Tumor-specific relative indications for amputation would include those lesions which could not be resected without destruction of neurovascular structures, infection in the surgical field, significant muscle or soft-tissue involvement and poor response to pre-operative chemotherapy.9
Van Nes rotationplasty is another option in the patient requiring a large resection of the distal femur.1,17,18,21 This procedure can also be performed when there is a required resection of the proximal tibia.43 The benefit of this procedure is that one can resect a large portion of distal femur or proximal tibia and still reconstruct a viable below“knee”amputation. One analysis of 30 rotationplasty procedures noted that there were 96.5% good and excellent results with a minimum follow-up of three years.24 Hanlon reported on 14 patients treated with rotationplasty for osteosarcoma followed for a mean of 8 years.18 All patients had good or excellent results according to the Enneking functional evaluation and all but one had no pain, with the other having only intermittent pain controlled by non-narcotic means. Thirteen of the fourteen patients had excellent emotional acceptance of the procedure, with the other having good acceptance. All patients reported that they would choose to have the procedure again.18
The reported complication rate with this procedure is relatively low when considering the patient population and surgical alternatives.17,18,24 Hanlon noted that three of 14 patients in his series required reoperations for surgical complications.18 However, a larger series noted that 30 of 70 patients had some type of early or late complication.17 In this series 16 of the early complications required operative intervention to manage their complication.17 The main complications described are vascular occlusions, nerve complications, non-union, infection and development of thigh length inequality.17,18,24
Despite the excellent results reported in the literature, the primary disadvantage of this technique is thought to be the cosmetic and potential psychological implications of this operation. The psychological disturbance often cited as a potential deterrent to this procedure has largely been discredited by the literature.21,33 In properly selected patients it is well-tolerated and has good results with relatively low incidence of immediate and long-term complications.18,20 Furthermore, although the ankle is subject to altered loading patterns the development of ankle arthrosis has not been shown in long-term radio-graphic or clinical follow-up.13
Along with improved treatment regiments, in recent years there has been an increasing focus on improving quality of life and function for the larger number of patients surviving their disease.20 Results of comparative studies differ with regard to whether amputation, rotationplasty or reconstruction offers the superior outcome.19,20,34 However, with improved device manufacturing and biomaterials there is an ever-gaining interest in limb-salvage by endoprosthetic reconstruction.9 In addition, there is a belief that limb salvage by prosthetic may offer psychological advantage due to an essentially normal outward appearing limb. Although it has been noted that most patients who survive this life-threatening disease adjust well to any necessary surgical treatment.38
Reconstructive options for salvage of the limb are many in the skeletally immature patient. They include allograft reconstruction, endoprosthetic reconstruction, and allograft-prosthetic composites. When accounting for the future growth potential of the pediatric patient, however, the options are more limited. Prosthetic reconstruction of the limb can be done with essentially two different modalities: modular prosthetic components and growing or expandable components. Modular components are thought to offer the advantage of increased prosthetic construct strength at maximal lengthening. The disadvantages are that each lengthening requires an operation that involves a rather large exposure. In the patient requiring multiple lengthenings this results in excessive scar formation and higher risk of infection.5 This option has never gained great favor in the young patient due to difficulties with participation in rehabilitation and the accompanying poorer results.41
Expandable components offer the advantage of either minimally invasive or non-invasive expansion procedures.6,11,12,22,32,35,37 The disadvantages of these devices, however, are the potential for failure of the expansion mechanism and failure of the prosthesis at maximal lengthening.12,22,23 The modular component may be more desirable in a patient where lengthening requirements will be minimal, whereas the expandable component may be desirable in a patient who will require multiple lengthening procedures.
In 1986 Lewis reported on six patients safely and successfully treated with implantation of an expandable and adjustable prosthesis.27 The expandable endoprosthesis is an option that offers the potential of limb salvage with equalized leg lengths at maturity. This can be accomplished through an expanding drive mechanism which can be accessed through a minimal approach or with newer prosthesis that are capable of non-invasive expansion in-situ. The advantages of limb salvage include potential for improved functional outcome. Rougraff and colleagues showed in their comparative series of 227 patients who underwent either limb salvage, above-the-knee amputation, or hip disarticulation that those undergoing limb salvage had significantly higher MSTS and Knee Society scores.34
The remainder of this article will focus on review of the expandable endoprosthesis for limb salvage of the pediatric malignancy.
When considering implantation of an expandable endoprosthesis wide excision should be the goal.9 Pre-operative planning will give an indication to the extent of the tumor and size of resection required.22 The amount of bone resected will limit the lengthening capability of the prosthesis; as the telescoping portion can be only as long as the prosthetic implanted.
An additional consideration is the acute lengthening. The benefits of acute lengthening are allowing for a longer rehabilitation period of the limb prior to undergoing the first lengthening procedure as well as possibly diminishing the number of future lengthening procedures needed. The amount of acute lengthening, however, must be conservative to avoid neurovascular complications. In general, it is recommended that this number not exceed 1-2 centimeters.36
Virtually all U.S. orthopaedic manufacturers offer modular knee systems which can be considered for the child at or near skeletal maturity. Decision regarding which device should be placed is based on surgeon preference as there is no data to guide selection of one particular manufacturer.
There are several manufacturers worldwide that produce expandable prostheses, they are mentioned in the following sections. In the United States, the only manufacturer that we are aware of is Wright Medical (Arlington, TN) who produces the Repiphysis system. We have had experience with this device at our institution.
When considering implanting an expanding endoprosthesis it is important to calculate estimated remaining growth.10 A child at or near skeletal maturity does not need an expandable device. Depending on their remaining growth the surgeon can consider an acute lengthening with implantation of a conventional endoprosthesis with or without a contralateral epiphysiodesis. Calculaion of remaining growth is instructive when counseling the patient and family members on surgical options and approximate number of lengthening procedures required.
Growth estimates are typically done utilizing data compiled by Anderson and Green.2 Using this data, any number of calculation schemes can be used to estimate growth anticipated in the contralateral limb and expected growth loss from resection of an involved physis or physes. One such option for a quick calculation is the Menelaus, or“arithmetic,”method.29 Using the assumption that on average the femoral physis grows 0.9 centimeters per year, and the tibial physis grows an average 0.6 centimeters, this allows the surgeon an estimate of anticipated discrepancy. It is important to note that chemotherapeutic and radiation treatment protocols have been shown to affect physeal growth, which further complicate these calculations.15,16
Dominkus and colleagues have evaluated the accuracy of growth prediction in children undergoing limb-sparing surgery for treatment of bone sarcomas.10 In their study they analyzed a series of fifteen patients with regard to their predicted discrepancy pre-operatively, and compared this to the amount of lengthening that was required by the time they reached skeletal maturity. They found that overall the elongation of the surgically treated limb exceeded the predicted growth by 24%.10 However, all patients in this study had limb lengths within 1 cm at skeletal maturity, which was due to the adaptability of the extendable device.
With use of the expandable prosthesis, pre-operative planning is paramount. Magnetic resonance imaging (MRI) can best define the extent of the tumor (Figure I).14 These should be done both prior to, and following neoadjuvant chemotherapy sessions. Planned resection and imaging data must be communicated to the manufacturer as these devices are custom made for each patient. Further more, careful examination of imaging studies will advise as to the proximity of the tumor to neurovascular structures. A prerequisite for limb salvage is the ability to maintain the neurovascular bundle without compromising local control of the malignancy.25 In addition, careful study of the tumor can advise toward the surgical approach made.22
Exposure of the tumor is done through an extensile approach as dictated by the pathoanatomy of the tumor. In general for tumors of the distal femur we prefer an anterior approach with a medial parapatellar arthrotomy Biopsy tracts are incorporated into the incision and el-lipsed out. The procedure begins with the resection of the tumor in the distal femur. Careful dissection is used to fully expose the tumor and any soft tissue extension (Figure 2). The collateral and cruciate ligaments are dissected off the tibia. The bone segment is measured and resected en bloc as planned based on pre-operative imaging. It is sent to pathology for confirmation of margins. After confirming margins the instruments are exchanged for clean tools and the surgical team applies new sterile gowns and gloves. The femoral osteotomy is then planed and the canal reamed to the pre-determined diameter. Both cemented and press-fit options are available.
The tibial cut is then made to resect the tibial spines and articular cartilage only. This tissue is sent for pathologic examination. The tibial canal is then reamed perpendicular to the physis to minimize trauma to the physis. A small cohort of patients having this device implanted has demonstrated that growth arrest does not occur following drilling and placement of a smooth press-fit stem across the tibial physis.31
Trial implants are then inserted and range of motion and stability in extension and flexion are tested. Once satisfied the final tibial component is placed. The femoral canal is then prepared with thorough irrigation. The femoral component is then placed with or without cement (Figure 3). Typically 1-2 cm of acute lengthening with the implant has been advocated to reduce the number of future lengthening procedures required.36 The hinge portion of the device is then assembled and the wound irrigated and closed. Post-operative radiographs are obtained to evaluate for implant and host bone complications (Figure 4).
The patient is seen at routine follow-up intervals. At each visit radiographs of the prosthesis are obtained to evaluate for complications. Orthoroentgenograms are regularly obtained to assess the limb length discrepancy. When a discrepancy exists of one to two centimeters, lengthening is undertaken.
The lengthening procedure is done under radiograph-ic guidance, typically in the fluoroscopy suite. Anesthesia is generally not required, and light intravenous sedation and/or analgesia are optional. In addition to negating the risk of the anesthetic to the patient, and additional advantage of performing awake lengthening is the ability to monitor pain and neurovascular examination.14
The locking mechanism of the prosthesis is located under fluoroscopy and the skin is marked overlying this. The electromagnetic coil is then applied around the leg at the location of the skin marking. The device is activated in 20 second intervals to achieve the desired lengthening. Images are examined after each lengthening interval.
Following the procedure the patient may require oral analgesics and assistive devices for a short period; however, this is quite variable between patients.
Due to the diminished strength of the expanded endoprosthesis, it has been recommended by some authors that children undergo conversion to a conventional fully-constrained hinge prosthesis at reaching skeletal maturity.7,35 However, there is no consensus on this point as some investigators feel that a fully expanded implant is strong enough to withhold stresses at full lengthening.36 In general there is increased concern for failure with full expansion of a telescoping prosthesis compared with a modular lengthened prosthesis.
Kenan reported on 54 children with osteosarcoma or Ewing's sarcoma who were treated with the Lewis Expandable Adjustable Prosthesis (LEAP).22 The LEAP prosthesis requires a small surgical approach to the chuck device for each lengthening. The authors noted that occasionally a fibrous membrane was found encasing the prosthesis which needed to be divided prior to lengthening. Of the 34 patients available for two to twelve-year follow-up, 24 patients had required revision procedures at some point, with all revisions successful. Twelve patients reached skeletal maturity without leg length discrepancy.
Schiller published on a group of six patients followed through skeletal maturity.35 His patients were treated with the Pafford-Lewis prosthesis and the Kotz Modular Femur Tibia Reconstruction system. Similar to the LEAP, this prosthesis expands telescopically requiring a small surgical approach to the site of the screw mechanism. These six patients had complications requiring a total seven revision procedures. On average each patient underwent 7.8 lengthening procedures for an average expansion length of 13.2 cm. Limb lengths at maturity were not reported in this series.
In 1993 Eckardt published a series of 12 patients treated with combination of different expandable pros-theses.12 They primarily used the LEAP in these patients, except for two patients who had Techmedica expandable prostheses. Ten were in the lower extremity, and two in the humerus. This series showed a high rate of failure of the implanted component, with seven of the 12 having a failure of the expansion mechanism. They largely converted to modular lengthening procedures in these patients. A second series was published by Eckardt in 2000 reporting on 32 patients treated with four different prosthetic styles.11 Twenty-two patients had a LEAP implanted, four Wright Medical modular prostheses, four Howmedica (Stryker/Howmedica/Osteonics; Rutherford, NJ) modular prostheses and two Techmedica prosthesis. In this series they continued to demonstrate a disappointing rate of device failure of 25%. However; the failures were largely salvageable and of the 19 surviving patients, limb salvage was successful in 16. Additionally, of the nine patients who reached skeletal maturity at the time of the report, six had equal leg lengths.
Schindler and colleagues evaluated their use of the Stanmore Mark II and Mark III‘growing prosthesis’(Stanmore Implants Worldwide, Stanmore, UK) in 18 children.37 This particular prosthesis was initially designed with ball-bearings (Mark II) and then with modular C-clips (Mark III) which could be implanted with subsequent surgeries to lengthen the prosthesis in a modular fashion. The next generation, the Mark IV, utilized the telescopic worm-wheel screw mechanism allowing for less invasive expansion procedures; however, these were not used in this particular cohort. Their series noted that at ten years there was a 100% failure rate of their prosthesis with amputation (two patients) or revision (ten patients) as the endpoint. Despite this, however, 84% of patients reported good or excellent results based on the MSTS scores. Further, function was estimated at 77% of expected normal function based on these scores. Patients had an average leg length discrepancy of 1.5 centimeters at skeletal maturity.
The Phenix (Phenix-medical, Paris, France) endoprosthesis was implanted in 15 patients as reported by Neel and colleagues.32 They noted patients had good early results with average Musculoskeletal Tumor Society (MSTS) scores of 90% at an average of 18 months. Of the three patients at skeletal maturity at the time of the report all had leg lengths within one centimeter. This is a custom manufactured device that utilizes stored energy in the form of a compressed spring for future lengthening. The energy is then selectively released in a closed fashion by subjecting the limb to an electromagnetic field which heats the polymer casing allowing for controlled expansion. The advantage of this technology is that lengthening is performed closed and, except for unusual circumstances, without general anesthesia. The Phenix is now manufactured as Repiphysis (Wright Medical Technology, Arlington, TN).
The only study, to our knowledge, examining the outcome of the Repiphysis prosthesis was reported by Gitelis et al.14 The fourteen patients in this series were reported at greater than two years follow-up. There was a 100% limb-salvage rate in this series, although there were five necessary revision surgeries all related to failure or fracture of the implanted components. Patients had an average MSTS score of 83.5% at the time of most recent follow-up.
A summary of the available literature is compiled in Table 1.
Aseptic loosening is the most frequently encountered complication in this particular limb-salvage procedure. This is largely due to the long lever arm created at the bone-implant interface which is necessitated by the extensive resection. Schindler noted that in their series the patients who developed aseptic loosening had an average of 57.8% of their femur length resected, compared with 46.2% in those who did not develop this complication. This was consistent with a finding reported by Cannon, who described a resection of greater than 60% increasing the likelihood of aseptic loosening.8
Mechanical failure is another frequently reported complication in the literature on this topic. This most frequently involved failure of the expansion mechanism, but there are also reports regarding fatigue fracture of the prosthesis.12,14,22 As experience with different expansion mechanisms has increased, this mode of failure seems to be reported with less frequency in the literature.
The series reviewed here indicate a variable incidence of deep infection. Kenan et al. reported four of 34 patients (11%) whereas many other series report no or only one of their cohort having an infection.22 These numbers are difficult to interpret given the inherent small numbers included in these studies. Deep infection with this procedure is a disastrous complication and many reports describe necessary treatment with limb ablation depending on the severity of the infection.
Less frequently reported complications include fracture, neurologic compromise and post-operative stiffness. These were infrequently and variably reported. A summary of reported complications is compiled in Table 1.
At this point the majority of the experience with expanding prostheses is represented in the distal femur and proximal tibia. This technology has been applied for the proximal femur,27 total femur27,37 and proximal humerus;12 but with even less experience. Further development and research is necessary to evaluate the utility of implants in these locations as well as others.
A common finding at time of lengthening or revision procedures reported in multiple publications was a pseu-docapsule surrounding the prosthesis which was stained black/gray by metallic wear debris.27,36 Schindler's group noted that this tissue had very low elasticity and required division to allow for lengthening. Improvements in implant design and biomaterials will improve wear characteristics and decrease complications related to lengthening and aseptic loosening.
An exciting new development in this field has been the creation of a“bioexpandable”endoprosthesis to address the issue of the increasing leverage between endoprosthetic/bone segments with subsequent lengthening.6,7 Baumgart and colleagues have reported on the MUTARS BioXpand device (Implantcast, Buxtehude) which employs the technique of callous distraction to lengthen the host bone segment. This device accomplishes this by use of the fully implantable motorized nail, the“Fitbone,”rather than external fixation devices as classically described by Ilizarov. There are, as of yet, no long term studies that we are aware of evaluating the outcomes and complications of this particular procedure.
The pediatric patient with a malignant bone tumor of the extremity is a challenging clinical scenario. Advances in adjuvant therapy have made limb-salvage procedures a realistic option without compromising survival or local recurrence. Surgical options to be discussed with the family include amputation, rotationplasty and prosthetic reconstruction. The advent of the expandable prosthesis has gained significant interest due to the appeal of limb salvage with a good cosmetic result and potential for equal limb length at skeletal maturity. Devices now exist which allow for non-invasive lengthening on an outpatient basis. Review of the literature demonstrates that this procedure has generally good patient reported outcomes but has a high complication rate, with aseptic loosening being the most common mode of failure. Further improvement in implant design and biomaterials may decrease these complications and work in these areas is underway. Currently, patients choosing this option should be counseled regarding the likelihood of future surgeries to manage the expected complications.