Limb salvage surgery has become the preferred option for the treatment of patients with primary bone sarcomas of the proximal tibia. When satisfactory surgical resection margins cannot be achieved, a traditional amputation or a modified amputation in the form of a Van Nes rotation- plasty may be possible to avoid the risk of local recurrence.11 However, for most patients with sarcomas of the proximal tibia, limb reconstruction is possible with various techniques. Knee arthrodesis is used infrequently because most patients favor having a mobile knee, which can be accomplished using allografts, allograft prosthetic composites, and custom or more commonly modular endopros- theses.1,2,4,7,16,29,32
Although endoprosthetic reconstruction after bone sarcoma resection has achieved excellent early to midterm results in the proximal and distal femur, the results after resection of the proximal tibia have been less successful. Inferior results associated with proximal tibia reconstruction are related to aseptic loosening, poor soft tissue coverage leading to an increased risk of infection, and disruption of the extensor mechanism affecting functional outcome.14,20,22,25,30,32 Most published results are based on cemented proximal tibia endoprostheses.7,14,17,22,27,30 Un- cemented tumor prostheses may be advantageous in the long term because of the potential for bone ingrowth and less risk of aseptic loosening.12,20,32,33
We determined the risk of aseptic loosening for patients who underwent resection of the proximal tibia for a primary bone tumor and reconstruction with an uncemented tumor endoprosthesis. We also examined whether additional complications influenced prosthetic failure, limb salvage and functional outcome.
MATERIALS AND METHODS
We identified 45 consecutive patients from our prospective database who underwent resection of a primary bone tumor of the proximal tibia between 1989 and 2003 and had the osteoarticular defect reconstructed with an uncemented proximal tibia tumor prosthesis. During this period, 24 additional patients were treated for a primary tumor of the proximal tibia with an allograft total knee arthroplasty composite (n = 5), a knee arthrodesis (n = 2), an intercalary allograft (n = 7), simple excision with no reconstruction (n = 2) for chondrosarcoma secondary to an underlying exostosis without intramedullary canal involvement, or an amputation (n = 8). Prospective database records and individual patient charts were reviewed to obtain general demographic, adjuvant and surgical data, local and systemic tumor relapse, and complications that could have an impact on prosthetic survival and limb salvage. There were 13 female and 32 male patients with an average age of 35 years (range, 14-73 years). The histologic diagnoses were osteosarcoma in 23 patients, chondrosarcoma in five, malignant fibrous histiocytoma of bone in five, giant cell tumor in four, leiomyosarcoma of bone in three, Ewing's sarcoma in two, and undifferentiated spindle cell sarcoma in two. One proximal tibia lesion was initially diagnosed and treated as an undifferentiated spindle cell sarcoma after an open biopsy but after resection was reclassified as a solitary adeno- carcinoma metastasis from an unknown primary tumor. Musculoskeletal Tumor Society staging at diagnosis identified four patients with Stage IB disease, three with Stage IIA disease, 34 with Stage IIB disease, and four with Stage III disease.9 Thirty- two of 35 patients with high-grade sarcomas other than chondrosarcoma received neoadjuvant chemotherapy. No patients received local radiotherapy. Minimum followup was 24 months unless the patient died. One patient with a history of psychiatric illness was lost to followup 7 months after treatment (at which time he was ambulating and free of disease) and was therefore excluded, leaving 44 patients as the focus of this study.
All reconstructions used the bone-ingrowth, fixed-hinge Kotz Modular Femur and Tibia Resection System (KMFTR; Howmedica, Stryker, Osteonics, Rutherford, NJ). The anchorage stem has a Madreporic porous coated surface with holes to allow for cross-screw fixation in the diaphysis through an attached side plate. The tibia diaphysis was underreamed 0.5 mm before insertion of the macroporous stem. If the stem could not be inserted, reaming was sequentially increased by 0.5 mm increments until a stable press fit was achieved, usually 0.5 to 1 mm larger than the diameter of the inserted stem. In the early part of the series, cross-screws were used for initial fixation in seven patients. We abandoned this after we determined the use of screws was not necessary for successful bone ingrowth and resulted in substantial stress shielding around the stem when compared with stems inserted without screws.5,12,21,32 The subsequent 37 patients had stems inserted without cross-screw fixation. This system includes modular components in various sizes designed to allow replacement of the amount of proximal tibia resected with a minimum resection length of 12 cm. The mean tibia resection length was 15 cm (range, 12-24 cm). The femoral component of the articulation is designed to preserve most of the native femoral condyles and is press fit into the intercondylar notch, also with an uncemented porous coated stem. However, in two patients, one with an intraarticular pathologic fracture and another with a skip metastasis in the distal femur, an extraarticular resection was performed that required an additional anchorage stem for the distal femur, and a distal femoral endopros- thesis for reconstruction. The prosthesis has a fixed hinge articulation that encompasses a polyethylene bearing and axle in a cobalt chrome cylinder.
Reconstruction of the extensor mechanism involved reattachment of the remnant patella tendon directly to the anterior tibia prosthesis using a screw/plate combination in 31 of 44 (70%) patients. When the remaining patella tendon was insufficient or absent after tumor resection, the position of the patella was maintained using nonabsorbable sutures and fascia lata (n = 3) or local tendon grafts (n = 2) attached to the anterior prosthesis. In two patients, the anterior tibia cortex/tuberosity was salvaged and this was secured to the tibia prosthesis with cables. One of these patients had a parosteal osteosarcoma involving the posterior tibia cortex, and the other had a proximal fibula osteosarcoma invading the posterior tibia. Local muscle flaps (n = 42) or a free flap (n = 1) were used to provide coverage for the prosthesis and reinforce the extensor mechanism repair.23 Nonabsorbable sutures were used to attach the muscle flap to the residual extensor mechanism. Only one patient had primary wound closure due to extensive skin devascularization after tumor resection and concern about the possibility of further skin loss if a local flap was elevated. This patient did not have a wound complication or deep infection develop. Postoperatively, patients were allowed touch weightbearing in a long leg cast with the knee hyperextended for 8 weeks if receiving no adjuvant treatment. To facilitate healing of the extensor mechanism repair and allow bone ingrowth onto the prosthesis in patients on chemotherapy we extended protected weightbearing to 12 weeks. If we had concerns about wound healing, we made a window in the cast or a removable splint was applied. After cast removal, patients began a physiotherapy program aimed at increased weightbearing, gait retraining, quadriceps strengthening, and gradual improvement in knee range of motion.
We identified complications that could affect prosthetic survival and limb salvage including aseptic loosening, prosthetic infection, fracture or instability, bushing or extensor mechanism failure, wound dehiscence, or local tumor relapse or vascular compromise. Prosthetic failure was defined as surgical removal of the original metallic prosthesis. Bushing failure that required only polyethylene bushing exchange was considered a minor complication,3,13,25,32,33 because the bone and prosthesis were not removed, the patient began to weight bear immediately following surgery without restrictions, and the hospital stay was brief. Patients underwent functional evaluation at a minimum 12 months after surgery using the Musculoskeletal Tumor Society (MSTS) 1987 and 1993 rating scales and the Toronto Extremity Salvage Score (TESS).6,8,10 Special attention was focused on function of the extensor mechanism by measuring knee range of motion, extension lag, and quadriceps strength based on the system of the Medical Research Council (MRC).24
Descriptive statistics were calculated using means, ranges, and proportions as appropriate. Actuarial rates for prosthetic survival and limb salvage and confidence intervals were calculated using the Kaplan-Meier method. Function for patients with no complications (n = 26) was compared with function (ie, TESS, MSTS 1987 and 1993) in patients who had additional surgery due to prosthetic, bushing, or extensor mechanism failure (n = 9).
At a mean final followup of 60 months (range, 9-152 months), 29 patients were alive with no evidence of disease, 13 died of their disease, one patient with known lung metastases died of a pulmonary embolus while being treated for a complication, and one patient died of an unrelated cancer. The mean survival was 74 months (range, 24-152 months) for patients who are still alive, and 27 months (range, 9-79 months) for those who died. Two of 44 (4.5%) patients had local tumor recurrences after margin-negative resections. One local recurrence occurred after extraarticular knee resection for a proximal tibia chondrosarcoma with an intraarticular pathologic fracture. The second local tumor relapse followed preoperative chemo- therapy and resection of a Ewing's sarcoma of the proximal tibia and skip metastasis of the distal femur. Both of these patients underwent amputations for local symptom control, but subsequently died of metastatic disease. Two patients had amputations because of vascular complications. One patient required posterior tibia vessel resection and reconstruction during the initial procedure because of vascular involvement by a large soft-tissue tumor mass. The vascular repair failed, necessitating an early above- knee amputation. Another patient had progressive and painful ischemia develop below the knee after multiple complications leading to an amputation 119 months after the initial procedure.
None of the 44 patients had aseptic loosening of the original uncemented tibia stem or the uncemented distal femoral component (Fig 1).
The most common complication, deep infection, occurred in seven of 44 (16%) patients and impacted outcomes. Six prosthetic infections occurred early after surgery, and five of these patients were receiving or had completed chemotherapy. Four cases of deep infection were treated with a two-stage revision procedure. In all four of these cases, the stem was solidly ingrown and was left in place during the revision procedures (Fig 2). Three patients underwent an amputation to control the infection. One infection occurred late after extensor mechanism revision surgery, one patient was severely malnourished because of chemotherapy treatment, and one patient was elderly and had poor soft tissue coverage, despite the use of a gastrocnemius flap. Three patients (7%) had superficial wound dehiscence postoperatively and were treated with débridement and reclosure or skin grafting without subsequent infection.
Mechanical complications affecting the prosthesis occurred in three of 44 (7%) patients and also influenced outcome. Two 13-mm tibia stems fractured through screw holes at 25 and 27 months after implantation. One stem fracture occurred in a patient with extensor mechanism failure. In both patients the stem below the fracture was solidly ingrown at revision surgery and was removed using trephines. Both patients underwent successful revision with a larger uncemented stem. One patient had rotational instability of the limb. A flange that secured the distal femur prosthesis to the base of the diaphyseal stem was broken and required exchange.
Three of 44 (7%) patients with polyethylene bushing failures presented clinically as having knee instability. All three bushing failures were treated with polyethylene bushing exchange at a mean 62 months (range, 50-76 months) after the original surgery. Two of the bushing failures seemed to be related to major trauma to the limb.
Failure of the extensor mechanism repair occurred in three of 44 (7%) patients, despite reinforcement of the original repair with gastrocnemius flaps. Extensor failures preventing active extension of the knee and large extensor lags occurred in two patients. In both of these patients, aggressive physiotherapy was initiated outside of our center less than 4 weeks after surgery despite instructions at discharge to delay range of motion exercises until 8 or 12 weeks postoperatively. One of these patients ambulated satisfactorily and required no additional treatment. The other patient underwent successful revision of the extensor mechanism, but had a late deep infection leading to amputation 61 months after the initial procedure. One patient, who had an extraarticular resection of the knee, including 12-cm segments of the proximal tibia and distal femur, dislocated the patella during postoperative physiotherapy and underwent successful extensor revision surgery. Another patient had persistent knee stiffness after postoperative immobilization, despite physiotherapy. The patient improved after manipulation of the knee under anesthesia 6 months after surgery.
The original implant was removed 14 times in 12 patients (27%) because of infection (n = 7), mechanical problems (n = 3; two stem fractures and one rotational instability), vascular complications (n = 2; one early and one late), and local tumor relapse (n = 2). If bushing failures are considered together with other complications that lead to prosthetic failure, 14 of 44 (32%) patients underwent additional surgery involving the prosthesis. Although four patients with prosthetic failure caused by infection and three patients with mechanical complications all had successful revision surgery, seven patients ultimately underwent an amputation, leaving 37 of 44 patients (84%) with successful limb salvage. The 5-year estimated rates of prosthetic survival, prosthetic survival without bushing revision, and limb salvage were 73% (95% confidence interval [CI] = 60%-87%; Fig 3), 70% (95% CI = 54%-85%), and 88% (95% CI = 78%-98%; Fig 4), respectively.
Functional outcome assessment was performed for 35 patients at a mean 3 years (range, 1-11 years) after the index procedure and revealed the following mean scores: TESS, 77/100 (range, 33-98); MSTS 1987, 25/35 (range, 13-31); and MSTS 1993, 75/100 (range, 33-97). The only patients not evaluated had early disease progression (n =3), early amputation (n = 3), or early prosthetic failure resulting in implant removal (n = 3). The mean knee flexion was 91° (range, 0°-110°). Twenty seven of 35 (77%) patients had flexion of 90o or greater. The mean knee extension lag was 6° (range, 0°-30°). Nineteen of 35 (54%) patients had no extension lag, five patients had an extension lag less than 5o, and another five patients had a lag of less than 10o. Quadriceps strength, as measured by the MRC grading system24 averaged 3.7/5 (range, 2-5). Only one patient had 5/5 quadriceps strength, whereas it was evaluated as 4+/5 in eight patients and 4/5 in 11 patients. Patients with quadriceps strength graded as 2/5 tended to have the most noticeable extensor lag.
The three functional outcome scores were similar for patients who had additional surgery for complications and those with no complications. However, patients who had additional surgery for complications exhibited greater extensor lags (mean 15.7 versus 3.9; p = 0.04), less quad- riceps strength (mean 3.1 versus 3.8; p = 0.01) and less knee flexion (mean 86o versus 92o; p = 0.6). After surgery for bushing failure (n = 3), prosthetic instability (n = 1), infection (n = 3) or stem fracture (n = 2), there was no significant change in any of the functional outcome measures. One patient who underwent major revision for a stem fracture showed improvement in all three function scores. The group of three patients with extensor mechanism failure tended to have lower functional scores and inferior functional measures (ie, extensor lag, strength, knee flexion) compared with patients with no complications, although inadequate power precluded a definitive conclusion.
Aseptic loosening of cemented proximal tibia tumor prostheses is the most common mode of failure, and has been documented to be as great as 86% at 5 years and 42% at 10 years.17,30 The uncemented macroporous KMFTR prosthesis we used in this study relies on a tight press fit of the distal femoral component and the proximal tibia diaphyseal stem to achieve initial stable fixation. The ingrowth potential of this prosthesis has been documented in other studies, most of which show low rates of aseptic loosening when used for distal femoral replacement.3,12,28,31-33 In this study, there were no cases of aseptic loosening of the original tibia diaphyseal stem or the distal femoral uncemented component. This supports the viability of uncemented fixation in proximal tibia endoprosthetic reconstruction and better medium-term loosening rates compared with some cemented devices. 14,17,18,20,22,27,30
The potential limitations of this study are the small number of patients, lack of controls, and short followup. However, all patients had been followed up prospectively from their initial presentation to our unit so that all aspects of their care and clinical data had been continually documented. Only one patient who was schizophrenic was lost to followup. Although this study included only 44 consecutively treated patients, it is the largest report of uncemented prostheses used for reconstruction at this anatomic site. There are few recent studies that focus exclusively on proximal tibia reconstruction and their mean followups were 59, 63, and 80 months compared to 60 months in this study.14,18,26
Fractures through screw holes in the diaphyseal stem are a major concern for this prosthesis. There were two(4.5%) tibia stem fractures in this study and in each case, the distal piece of the stem below the break was solidly ingrown and was removed at revision surgery with difficulty and additional bone loss. Other studies reported stem fractures in as many as 10% of these implants.3,12,25,31-33 Stem fractures have also been reported with solid stems without screw holes, but are not as common.14,20,22,27,30 However, one study of 133 cemented proximal tibia custom endoprostheses reported fractures of 12 (9%) stems.26 Because we no longer use cross-screws for initial fixation due to concerns related to stress shielding, simply having the manufacturer produce a similar but solid uncemented stem would likely minimize this complication.5,12,21,32
We now use a new version of the distal femoral KMFTR prosthesis with a rotating-hinge knee mechanism, although it is not yet available for uncemented proximal tibia implants. This new joint design will likely help offset some of the stresses on the prosthesis and could diminish the risk of stem breakage and polyethylene bushing failure, which occurred in three of the 44 patients (7%) in this study. The failed bushings had thin polyethylene flanges typical of an early version of this prosthesis, and were previously reported to fail at a rate of 41% (39/95) at 5-year followup.3 Subsequent modifications to the bushings have made failures less common.
Infection is a devastating complication of endoprosthetic reconstruction. Prosthetic infections occurred in seven of 44 patients (16%) in this study and three of these ultimately underwent amputations. Prosthetic infection is especially problematic in the proximal tibia because soft tissue coverage usually is suboptimal after sarcoma resection. Although all but one of our patients had muscle flaps as part of their reconstructions, the infection rate was still higher than that typical of other lower limb sites.11,12,14,18,19,22,23,32 Three other studies that used the same KMFTR prosthesis, although predominantly for distal femur reconstructions, also reported high rates of infection, ranging from 10% to 15%.3,25,33 In comparison, two recent studies of cemented proximal tibia endopros- theses reported infections in four of 13 patients (31%)22 and 18 of 151 (12%)14 patients. We previously found many patients with prosthetic infections have identifiable predisposing factors.12 Consistent with this, neutropenia, malnutrition from chemotherapy, old age, and poor soft tissues likely contributed to deep infection in six of seven patients in this study. Whether the porous surface of this particular stem has a role in the high infection rate reported in this study remains to be determined.
In patients with an infected but well ingrown tibia diaphyseal anchorage piece and distal femoral component, we now retain the stems and replace the modular components with an antibiotic spacer as part of a staged revision protocol (Fig. 2). This method was used to treat four prosthetic infections in this study and is a major advantage compared with two-stage revisions for cemented megaprostheses in which the stem and intramedullary cement typically must be removed.13 One-stage revision without exchange of an ingrown anchorage stem has also been reported for uncemented KMFTR infections.15
The inferior functional outcome for patients with proximal tibia endoprostheses compared with distal femoral reconstructions is directly related to disruption of the extensor mechanism. We were able to reattach the remnant patella tendon directly to the prosthesis in 31of 44 patients (70%) and the extensor repair was reinforced with a local rotational flap or free flap in 43 of 44 patients. This method of repair resulted in an average knee extension lag of 6°, which compares favorably with two other studies that reported extensor lags of 10° and 30° after proximal tibia replacement.14,20 In comparison, the mean active knee flexion of 91° achieved by our patients was less than reports of 104° and 110° in the same two studies.14,20 This improvement in extensor lag at the cost of decreased knee range of motion could be partially related to the prolonged period of knee immobilization in hyperextension for our patients following surgery, though the maximum possible flexion obtainable is limited in part by the fixed hinge design of this implant.3,12,25,32,33 Quadriceps strength of3.7/5 in our patients based on MRC evaluation was almost identical to that reported by Grimer et al.14 The MSTS functional scores in this study were also similar to other reports,14,22 but lower than the TESS scores reported by our patients. This difference in functional outcome measures reflects the fact that MSTS is completed by the clinician and evaluates clinical parameters mainly at the anatomic level, whereas TESS is completed by the patient and reflects their ability to perform activities of daily living to which they typically adapt after surgery.
Postoperative complications, particularly those having additional surgery for revision, may have an impact on patient function. There were no significant differences in the three functional outcome scores between patients with no complications and those having revision procedures, although the objective functional parameters of knee flexion, extensor lag, and quadriceps strength all were inferior after subsequent surgery for complications. However, the three of 44 patients (7%) who had extensor mechanism failure had lower functional scores and inferior flexion, extensor lag, and strength, thereby emphasizing the importance of extensor repair at the initial surgery and protection of the repair postoperatively.
It has been well documented that patients who require a prosthesis for reconstruction after a tumor of the proximal tibia have a higher rate of complications and failure than for other lower extremity sites.12,14,17,20,22,25,26,30,32 Although 12 of 44 patients (27%) in this study underwent additional surgery to manage complications, most prosthetic-related problems were successfully revised, leading to limb salvage in 37 of 44 patients (84%). The estimated 5-year prosthetic survival rate of 73% in our patients is higher than other studies, which reported survival of 45% to 60% for cemented tibia endoprostheses at 5 years.14,17,20,22 With continued followup beyond 5 years there seems to be an ongoing need for revision surgery of cemented tibia endoprostheses. In comparison, the 5-year limb salvage rate of 88% in our study compared favorably with 85.5% in a series by Natarajan et al26 and was slightly higher than other reports of 78%,14,17 although these figures remained more stable over time. The low rate of local tumor relapse in this study (2/44, 4.5%) also compares favorably with other series in which local recurrence has been as great as 11%. This low rate also attests to the safety of limb salvage for sarcomas of the proximal tibia.7,14,18,20,25,28,31-33
This study shows the advantages of uncemented proximal tibia endoprosthetic fixation compared with comparable cemented devices in terms of aseptic loosening and revision for infection at midterm followup. Despite the potential obstacles for the patient and the surgeon to achieve successful limb salvage for tumors at this location, endoprosthetic replacement remains a valuable and favored reconstructive option.
1. Brien EW, Terek RM, Healey JH, Lane JM. Allograft reconstruction after proximal tibial resection for bone tumors. An analysis of function and outcome comparing allograft and prosthetic reconstructions. Clin Orthop Relat Res
2. Campanacci M, Costa P. Total resection of distal femur or proximal tibia for bone tumours. Autogenous bone grafts and arthrodesis in twenty-six cases. J Bone Joint Surg Br
3. Capanna R, Morris HG, Campanacci D. Del Ben M, Campanacci M. Modular uncemented prosthetic reconstruction after resection of tumours of the distal femur. J Bone Joint Surg Br
. 1994;76: 178-186.
4. Clohisy DR, Mankin HJ. Osteoarticular allografts for reconstruction after resection of a musculoskeletal tumor in the proximal end of the tibia. J Bone Joint Surg Am
5. Davis AM, Damani M, White LM, Wunder JS, Griffin AM, Bell RS. Periprosthetic bone remodeling around a prosthesis for distal femoral tumors: longitudinal follow-up. J Arthroplasty
. 2005;20: 219-224.
6. Davis AM, Wright JG, Williams JI, Bombardier C, Griffin A, Bell RS. Development of a measure of physical function for patients with bone and soft tissue sarcoma. Qual Life Res
7. Eckardt JJ, Matthews JG 2nd, Eilber FR. Endoprosthetic reconstruction after bone tumor resections of the proximal tibia. Orthop Clin North Am
8. Enneking W. Modification of the system for functional evaluation of the surgical management of musculoskeletal tumors
. New York, NY: Churchill Livingstone; 1987.
9. Enneking WF. A system of staging musculoskeletal neoplasms. Clin Orthop Relat Res
10. Enneking WF, Dunham W, Gebhardt MC, Malawar M, Pritchard DJ. A system for the functional evaluation of reconstructive procedures after surgical treatment of tumors of the musculoskeletal system. Clin Orthop Relat Res
11. Gottsauner-Wolf F, Kotz R, Knahr K, Kristen H, Ritschl P, Salzer M. Rotationplasty for limb salvage in the treatment of malignant tumors at the knee. A follow-up study of seventy patients. J Bone Joint Surg Am
12. Griffin AM, Parsons JA, Davis AM, Bell RS, Wunder JS. Uncemented tumor endoprostheses at the knee: root causes of failure. Clin Orthop Relat Res
13. Grimer RJ, Belthur M, Chandrasekar C, Carter SR, Tillman RM. Two-stage revision for infected endoprostheses used in tumor surgery. Clin Orthop Relat Res
14. Grimer RJ, Carter SR, Tillman RM, Sneath RS, Walker PS, Unwin PS, Shewell PC. Endoprosthetic replacement of the proximal tibia. J Bone Joint Surg Br
15. Holzer G, Windhager R, Kotz R. One-stage revision surgery for infected megaprostheses. J Bone Joint Surg Br
16. Hornicek FJ Jr, Mnaymneh W, Lackman RD, Exner GU, Malinin TI. Limb salvage with osteoarticular allografts after resection of proximal tibia bone tumors. Clin Orthop Relat Res
. 1998;352: 179-186.
17. Horowitz SM, Glasser DB, Lane JM, Healey JH. Prosthetic and extremity survivorship after limb salvage for sarcoma: how long do the reconstructions last? Clin Orthop Relat Res
18. Horowitz SM, Lane JM, Otis JC, Healey JH. Prosthetic arthroplasty of the knee after resection of a sarcoma in the proximal end of the tibia. A report of sixteen cases. J Bone Joint Surg Am
. 1991;73: 286-293.
19. Jeys LM, Grimer RJ, Carter SR, Tillman RM. Risk of amputation following limb salvage surgery with endoprosthetic replacement in a consecutive series of 1261 patients. Int Orthop
20. Kawai A, Healey JH, Boland PJ, Athanasian EA, Jeon DG. A rotating-hinge knee replacement for malignant tumors of the femur and tibia. J Arthroplasty
21. Lan F, Wunder JS, Griffin AM, Davis AM, Bell RS, White LM, Ichise M, Cole W. Periprosthetic bone remodelling around a pros- thesis for distal femoral tumours. Measurement by dual-energy X-ray absorptiometry (DEXA). J Bone Joint Surg Br
. 2000;82: 120-125.
22. Malawer MM, Chou LB. Prosthetic survival and clinical results with use of large-segment replacements in the treatment of high- grade bone sarcomas. J Bone Joint Surg Am
23. Malawer MM, McHale KA. Limb-sparing surgery for high-grade malignant tumors of the proximal tibia. Surgical technique and a method of extensor mechanism reconstruction. Clin Orthop Relat Res
24. Medical Research Council. Aids to the investigation of the peripheral nervous system
. London, England: Her Majesty's Stationary Office; 1976.
25. Mittermayer F, Krepler P, Dominkus M, Schwameis E, Sluga M, Heinzl H, Kotz R. Long-term follow-up of uncemented tumor endoprostheses for the lower extremity. Clin Orthop Relat Res
. 2001; 388:167-177.
26. Natarajan MV, Sivaseelam A, Rajkumar G, Hussain SH. Custom megaprosthetic replacement for proximal tibial tumours. Int Orthop
27. Roberts P, Chan D, Grimer RJ, Sneath RS, Scales JT. Prosthetic replacement of the distal femur for primary bone tumours. J Bone Joint Surg Br
28. Sanjay BK, Moreau PG. Limb salvage surgery in bone tumour with modular endoprosthesis. Int Orthop
29. Sim FH, Beauchamp CP, Chao EY. Reconstruction of musculoskeletal defects about the knee for tumor. Clin Orthop Relat Res
. 1987; 221:188-201.
30. Unwin PS, Cannon SR, Grimer RJ, Kemp HB, Sneath RS, Walker PS. Aseptic loosening in cemented custom-made prosthetic replacements for bone tumours of the lower limb. J Bone Joint Surg Br
31. Witt JD, Marsden FW. The functional evaluation of patients with primary malignant tumours about the knee treated by modular endoprosthetic replacement. Aust N Z J Surg
32. Wunder JS, Leitch K, Griffin AM, Davis AM, Bell RS. Comparison of two methods of reconstruction for primary malignant tumors at the knee: a sequential cohort study. J Surg Oncol
33. Zwart HJ, Taminiau AH, Schimmel JW, van Horn JR. Kotz modular femur and tibia replacement. 28 tumor cases followed for 3 (1-8) years. Acta Orthop Scand