Application of artificial prosthesis reconstruction techniques in malignant tumors around the knee joint

Introduction

About 30% of all malignant bone tumors occur in the distal femur and proximal tibia, and 70% of osteosarcomas, the commonest primary malignant bone tumor, occur around the knee joint ¹ . In the last 30 years, with the progress and development of imaging techniques, combination therapy and surgical techniques, limb salvage surgery has gradually replaced amputation and become the primary surgical method for managing malignant bone tumors in the extremities.

There are many reconstruction methods for bone defects after resection of malignant bone tumors around the knee joint, namely: (i) vascularized fibular grafts; (ii) osteoarticular allografts; (iii) inactivation and reimplantation; (iv) artificial prosthesis replacement; (v) allograft‐prosthesis compound reconstruction. The decision as to which reconstruction technique to use depends on the nature of the tumor, its anatomical site, extent of invasion, size of bone defect and clinical condition of the patient.

The Massachusetts General Hospital has done extensive research on the application of massive bone allograft over a long period, and in 1996 reported 718 cases of bone allograft after resection of bone tumors in the extremities and reconstruction (including 386 cases of osteoarticular allograft). Their postoperative fracture rate was 19%, nonunion 17%, infection 11% and instability 6% ² . In 2004, they reported the evolution of reconstruction techniques for limb salvage following resection of bone tumors during the eight years between 1996 to 2004: the number of artificial prosthesis reconstructions had increased by 40%, while bone allografts had decreased by 50% over the same period. These differences can be explained by the following: for bone allografts the non‐weight‐bearing period is longer and nonunion and infection rates higher; whereas artificial prosthesis reconstruction has the advantages of early weight‐bearing, good stability and few early complications ³ . Nowadays, artificial prosthesis reconstruction has become a common reconstruction method for bone defects after resection of malignant tumors around the knee joint.

Artificial prosthesis replacement of distal femur tumors

Exposure and resection

For resections of distal femur tumors, medial or anterior incisions can be selected. The former is more convenient, affording easier exposure and protection of the femoral artery and vein and causing minor damage to the knee extensor mechanism; in addition gastrocnemius flap transplantation may be achieved by extending the incision distally if necessary. The medial incision starts from the medial midpoint of the thigh and crosses the knee joint beside the medial patella to arrive at the distal tibial tuberosity (Fig. 1). An oval incision should be made around the biopsy scar, so that both the biopsy track and the whole tumor can be resected. A medial incision enables wide exposure of the distal half of the femur, sartorius muscle, knee joint, popliteal fossa, and proximal half of the tibia (Fig. 2). In order to protect the popliteal vessels and sciatic nerve, the medial hamstring should be separated and retracted to expose the popliteal fossa. Several joint branch vessels of the popliteal artery and vein should be ligated or excised to establish a space between the back of the femur and the popliteal vessels. The distal femur in the intermuscular space between the rectus femoris and vastus medialis should be exposed, retaining all of the vastus intermedius on the distal femur and part of the vastus medialis on the tumor. In some cases, part of the vastus lateralis should also be retained on the tumor, thus preserving a cuff of normal tissue around it in order to achieve wide resection (Fig. 3). The knee joint should be incised longitudinally along the medial patella and the cruciate ligaments divided to completely expose the articular cavity.

Figure 1.

The line of medial incision for distal femur bone tumor resection.

Figure 2.

The medial incision enables wide exposure of the distal ½ of the femur, sartorius muscle, knee, popliteal fossa and proximal ½ of the tibia.

Figure 3.

The bone tumor is exposed, together with a cuff of normal tissue surrounding the tumor tissue.

These days, MRI has replaced plain radiography and become a major foundation for determining the osteotomy plane. Compared to plain radiographs, osteotomy planes determined according to MRI reduce the length of the osteotomy to less than 2 cm on average. The MRI‐identified osteotomy plane is where the normal marrow signal on T1WI changes to an abnormal signal. Marking the locations of the proximal and distal ends of the osteotomy on the bone surface before osteotomy is helpful in adjusting the femoral rotary line of force of reconstruction. The branches of the popliteal vessel should be ligated, the cruciate ligaments divided and the tumor‐affected bone completely removed on the basis provided by the preoperative MRI (Fig. 4). The neoplastic bone should be removed from the surgical field and about 1 cm of tibial plateau bone resected. The osteotomy plane should be perpendicular to the long axis of the tibia.

Figure 4.

The branches of the popliteal vessel are ligated, the cruciate ligaments severed and the tumor bone completely removed according to the preoperative MRI.

Prosthesis installation

Over recent years, the application of modular prostheses has gradually increased. The length of neoplastic bone resected should be measured and, in adults, a prosthesis of equal length used, whereas in children, the prosthesis should be 1 cm longer than the neoplastic bone removed, in order to compensate for lengthening due to epiphysis growth in the future. The diameter of pulp chamber file used to expand the marrow cavity of the distal femur or proximal tibia should be 2 mm larger than that of the prosthesis stem selected.

Before immobilizing with bone cement, a trial prosthesis of appropriate size should be implanted, the movement track of the joint checked and the force line of the prosthesis defined and adjusted to prevent prosthesis intorsion or extorsion, which may otherwise compromise knee joint function and the activity track of the patella. The marrow cavity should be thoroughly rinsed and a bone cement plug implanted in the cavity. After injecting bone cement into the bone cavity with a bone cement gun, the distal and proximal prosthesis handles should be implanted and all components connected, including the femoral marrow cavity handle, femoral body (extended section), femoral condyle, axletree, polyethylene liner and tibial weight‐bearing support (Fig. 5). Care should be taken to ensure that the tibial and femoral components are in good alignment and apposition according to the intertrochanteric line, tibial tuberosity and marking lines previously drawn. Patellar replacement is usually not performed in patients with obvious patellar degeneration. A typical case that has been reconstructed with a modular prosthesis is shown in Fig. 6.

Figure 5.

Modular prostheses are used for reconstruction of the distal femur after tumor resection.

Figure 6.

Osteosarcoma of the distal femur in a 16‐year‐old boy. (a) Preoperative AP and lateral views of the distal femur show an osteolytic lesion, large bone defects and involvement of the epiphysis. (b) MRI shows a large volume tumor before administration of neoadjuvant chemotherapy (c) MRI after neoadjuvant chemotherapy shows the tumor volume is smaller than before and necrosis of the tumor is evident (d) Radiographs after modular prosthesis reconstruction show alignment of the prosthesis is satisfactory.

Soft tissue reconstruction

After completion of prosthesis installation, care should be taken to ensure that peripheral soft tissue completely covers the prosthesis. The residual vastus medialis can be sutured to the rectus femoris, and the sartorius freed forward and sutured to the peripheral soft tissue to increase the coverage of medial soft tissue. If the soft tissue defect is too large to cover by the above procedure, rotation of the medial head of the gastrocnemius to cover the prosthesis can be considered.

Postoperative treatment

The affected limb should be elevated for 3 days, a negative pressure drainage tube kept in place for 3–5 days, and intravenous antibiotics stopped after extraction of the drainage tube. A splint to limit joint movement should be worn for 2–3 weeks to facilitate soft tissue union and functional recovery of the extensor mechanism. When the muscle strength of the quadriceps femoris has recovered, the patient can begin functional exercise and gradually resume weight bearing with the help of a support.

Artificial prosthesis replacement of proximal tibia tumors

Exposure and resection

In resection of proximal tibial tumors, an anterior‐medial incision is selected, starting from the medial distal third of the femur and crossing the parapatellar area downwards along the medial calf to the distal third of the tibia. This incision easily exposes the medial head of the gastrocnemius and allows cover of the prosthesis and patellar ligament insertion by rotation. Resection of the biopsy track is achieved as described above. According to the tumor resection principles proposed by Enneking et al., the margin around the tumor should be wide or narrow ⁴ . The medial and lateral skin flaps should be separated from the underlying fascia. The skin flap should be severed 3 cms proximal to the medial hamstring insertion, the medial head of the gastrocnemius freed, the soleus split to expose the popliteal fossa vessels and the medial sural artery retained. The cruciate ligaments should be divided close to the femur. Part of the tibialis anterior, popliteus and soleus muscles should be retained on the tumor. The common peroneal nerve should be divided and protected and the biceps femoris divided 2 cm from its insertion. If the tumor involves the tibiofibular joint, the popliteal artery should be retracted posteriorly; the anterior tibial vessels separated and ligated at the inferior border of the popliteus and the fibula divided 8 cm from its head. When performing extra‐articular resection, there is no need to incise the knee joint capsule. The femur should be divided 2 cm above the knee joint capsule, the patella split in the coronal plane and the patellar ligament separated from the infrapatellar fat pad; the remaining steps are as same as for internal joint resection. An osteotomy of the distal tibia should be performed according to the osteotomy plane determined by MRI preoperatively.

Prosthesis installation

The diameter of tibial pulp chamber file should be 2 mm larger than that of the selected prosthesis stem. The marrow cavity should be thoroughly irrigated to facilitate bone cement fixation. An appropriate trial prosthesis should be selected according to the length of neoplastic bone resected and the diameter of the marrow cavity and the joint movement track, force line, length of prosthesis, joint movement and stability checked. If satisfactory, the osteotomy plane should be marked before removing the trial prosthesis to prevent rotation when installing the permanent prosthesis. After implanting femur and tibia marrow cavity plugs, bone cement should be injected and the prosthesis components installed.

Soft tissue reconstruction

The patellar ligament should be threaded with nonabsorbable suture material, immobilized firmly at the front ring or hole of the proximal tibial prosthesis and the knee joint allowed to flex to 30°–40°. The medial head of the divided gastrocnemius should be rotated to cover the patellar ligament insertion and its upper margin sutured to the medial and lateral joint capsule and its lateral margin to the deep fascia anterior to the tibia and fascia of the tibialis anterior to make it possible for the gastrocnemius to completely cover the prosthesis and to increase the strength of the knee extensors post‐reconstruction (Fig. 7).

Figure 7.

Schematic diagrams depicting transference of thep medial gastrocnemius to reconstruct the extensor mechanism.

Postoperative treatment

The afflicted limb should be raised for about 5 days, with negative‐pressure drainage in place for 3–5 days and an intravenous antibiotic drip until after removal of the drainage tube. The knee joint should be immobilized for 4–6 weeks with support to facilitate union of the ligamentum patellae and surrounding soft tissue. Then exercise flexion and extension of knee joint can be commenced with support, the range of motion being limited to 0°–30°.

Prosthesis survival rate and factors influencing it

The main issue that concerns doctors and patients is the useful life of the prosthesis. In contrast to biological reconstruction, complications of artificial prosthesis reconstruction tend to increase gradually with time and some patients, especially young ones, may require prosthesis replacement. After routine artificial prosthesis replacement of the knee joint, the five year prosthesis survival rate (PSR) is above 95%; however, after reconstruction for malignant bone tumors, the PSR is less, having been reported to be 72%–93%, 53%–86% and 48%–66% at five, ten and fifteen years postoperatively, respectively ⁵ , ⁶ , ⁷ , ⁸ . The PSR is influenced by many factors, including the prosthesis implantation site, length of bone resected, amount of surrounding soft tissue resected, prosthesis design and amount and type of activity.

Sites

For prosthesis reconstruction of the distal femur, Kawai et al. reported 55 cases using Lane‐Burstein hinged prosthesis reconstruction; after three, five and ten years their PSRs, without replacement of any component, were 85%, 67% and 48%, respectively ⁶ . For artificial prosthesis reconstruction of the proximal tibia, Wu et al. reviewed 44 cases, their PSRs after five and ten years were only 72.3% and 53.1%, respectively ⁷ . Unwin et al. reported 1001 cases using Stanmore customized prosthesis reconstruction, among which there were 493 tumors of the distal femur, 263 of the proximal femur and 245 of the proximal tibia ⁸ . After ten years, the PSRs for the proximal femur were 93.8%, distal femur 67.4% and proximal tibia 58%. The above study shows the PSR is better for the distal than the proximal tibia.

Length of bone resected and amount of surrounding soft tissue resected

Kawai et al. reported 82 cases of artificial prosthesis reconstruction after resection of distal femur malignant bone tumors ⁹ . The PSR of the group in whom <40% of the distal femur length was resected was considerably better than that of the group with resection length ≥40%. Unwin et al. analyzed the PSR with three different lengths of proximal tibia resected (<40%, 40%–60%, >60%) ⁸ . The group with resection length <40% had the highest PSR, followed by the 40%–60% group, the >60% group having the smallest PSR. It is thus clear that the longer the neoplastic bone resected, the worse is the PSR.

The greater the extent of soft tissue surrounding the tumor resected, the worse is the PSR. After wide resection of the quadriceps femoris, its buffer action on limb torsional forces decreases. The increase in torque of the prosthesis axis and the weakening of the quadriceps femoris contraction force can cause aseptic prosthesis loosening, which can reduce the PSR. Kawai et al. analyzed the PSR in two groups, one with intact quadriceps femoris and the other with complete resection of the quadriceps femoris; the latter's PSR was considerably decreased ⁹ .

Prosthesis design

Artificial prostheses are classified into customized, modular and adjustable prostheses according to design types, and into hinged and rotary hinged prostheses according to sports mechanics. Wu et al. reported artificial prosthesis reconstruction after proximal tibia malignant bone tumor resection, including 9 cases of customized and 35 of modular prostheses ⁷ . The PSRs after five and ten years were 44.4% and 22.2%, respectively, for the former, and 81.4% and 65.3%, respectively, for the latter. There was a significant difference between the two groups. Schwartz et al. reported 23 cases of customized and 29 of modular prostheses; their PSRs after 15 years were 63% and 88%, respectively ¹⁰ . Kawai et al. reported 51 cases of Lane‐Burstein hinged prostheses with PSR after 5 years of 64%, and 31 cases of Finn rotary hinged prostheses with PSR after 5 years of 82% ⁹ . The above studies show that the survival rates for modular and rotary hinged prostheses are better than for customized and Lane‐Burstein hinged prostheses.

Therefore complete tumor resection, maximal protection of surrounding soft issue and appropriate choice of artificial prosthesis type are all associated with increased PSR after reconstruction following resection of malignant bone tumors surrounding the knee joint.

Diagnosis and treatment of prosthesis infection

Due to the specificity of tumor surgery, the periprosthesis infection rate is relatively high after reconstruction following resection of malignant tumors surrounding the knee joint. The incidence rate has been reported to be between 3.6% and 37.5% (10%–20% in most studies); thus this serious complication is secondary only to local recurrence. If such infections are inappropriately treated, the amputation rate can vary between 37% and 87% ⁷ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ . Reasons for periprosthesis infection include: (i) the periprosthesis area may lack effective soft tissue coverage after reconstruction involving substantial resection of soft tissue, especially with proximal tibia prostheses; (ii) poor patient immune status due to neoadjuvant chemotherapy or radiotherapy; and (iii) long operation times and wide exposure increase the risk of infection.

Many scholars have studied the infection rate after prosthesis reconstruction of tumors surrounding the knee joint. Jeys et al. followed up 1264 cases of tumor prostheses for an average of 5.8 years, and found the infection rate for the proximal tibia was the highest, namely 23.1%. For the distal femur it was 10.3%, and for the humerus the lowest, at 1.1% ⁶ . Flint et al. reported an infection rate for the proximal tibia of 15.9% ¹⁷ . Sharma et al. reported an infection rate for the distal femur of 7.8% ¹⁸ . It is thus clear that the infection rate after prosthesis reconstruction is higher for the proximal tibia. A study by Jeys et al. showed the most common pathogen is the coagulase negative staphylococcus, a finding which can provide guidance in selecting antibiotics for these infections ¹⁶ .

Diagnosis

Periprosthesis infection is a disastrous complication. The highest incidence of periprosthesis infection is within two years of initial prosthesis reconstruction or one year of prosthesis revision, therefore a high index of suspicion for this complication is appropriate during these periods. The diagnosis of periprosthesis infection depends mainly on clinical symptoms, combined with laboratory, including microbiological, tests. The symptoms include fever, local swelling, increased skin temperature, obvious nocturnal pain, rest pain, sudden otherwise unexplained pain developing in a knee joint whose function has recovered and deterioration in joint function. Although C‐reactive protein concentration, erythrocyte sedimentation rate, white blood cell count and other inflammatory markers lack specificity, they can offer guidance. If the postoperative erythrocyte sedimentation rate and C‐reactive protein concentration increase again after having returned to normal, or do not return to normal at all, periprosthesis infection should be suspected. Bacterial culture of fluid obtained by paracentesis from the diseased site is the most reliable method for diagnosing periprosthesis infection.

Treatment

Since these infections occur at the implant‐bone interface, complete cure is difficult to achieve without removing the prosthesis and cement. The success rate for single antibiotics is very low. Accurate assessment of the indications is very important for determining whether to use single‐ or multi‐phase arthroplasty for infections surrounding prostheses.

Single‐phase arthroplasty is recommended in the following circumstances: infection with Gram‐positive bacteria of low virulence; when antibiotic‐loaded bone cement was used in the initial prosthesis arthroplasty; and where long‐term postoperative antibiotics are used. Many scholars in Europe have proposed implanting new prostheses, using gentamicin‐loaded bone cement, at an early stage after complete debridement, lavage and removal of the original implant ¹⁹ . However, it has been reported that infection control and limb‐salvage rates are higher after staged arthroplasty; accordingly, many scholars support staged arthroplasty.

The indications for staged arthroplasty are as follows: inability to achieve complete debridement, a non‐bone cement prosthesis having been used in the initial arthroplasty, inconclusive results of bacterial culture, mixed infection, Gram‐negative bacterial infection and chronic sinus tract formation. The key steps of staged arthroplasty are: (1) removal of the artificial prosthesis; (2) complete debridement, including infected soft tissue surrounding the prosthesis, scar tissue and intramedullary bone cement; (3) implantation of a spacer. The size of the bone cement spacer should match that of the prosthesis to be reimplanted. An appropriate antibiotic determined by drug sensitivity tests should be added to the bone cement; and (4) intravenous antibiotics for at least 6 weeks. After implantation of the spacer, further bacterial cultures will determine whether there is persistent infection. If the cultures are positive, the initial spacer should be removed and replaced by a new one and the antibiotic changed, until bacterial cultures are negative; (5) reimplantation of the prosthesis after completely control of infection. The interval between the two operations of staged arthroplasty should be at least 6–8 weeks apart; generally six months or so. We recommend that timing of prosthesis reimplantation should be based mainly on the patient's general condition, the extent of preoperative infection, adequacy of infection control, state of previous debridement and the condition of the local soft tissue. If the pathogenic bacteria are virulent Gram‐negative bacteria, there is serious mixed infection or the patient is in a poor general condition, the time to prosthesis implantation should be extended as long as possible ²⁰ .

Grimer et al. reported 34 cases of periprosthesis infection treated by staged arthroplasty ¹⁹ . Their infection control rates after one, five and ten years were 91%, 74% and 65%, respectively, and the final limb‐salvage rate was 82%. Manoso et al. reported 11 cases of staged arthroplasty for infection after knee joint salvage surgery ²¹ . They reimplanted the artifical prosthesis after an average of three debridements and covered the prosthesis with freed latissimus dorsi or rectus abdominis flaps; their final limb‐salvage rate was 100%. This study suggests that debridement is a very important component of staged arthroplasty for infection of post‐tumor artificial prostheses. Generally, a single debridement is insufficient. Multiple debridements may assure achievement of negative local bacterial cultures, which is the precondition for successful reimplantation of the prosthesis.

Influential factors and treatment of aseptic loosening of artificial prostheses following tumor resection

After artificial prosthesis reconstruction following resection of malignant tumors around the knee joint, the rate of aseptic loosening of the prosthesis varies considerably according to the prosthesis implant site, prosthesis design, length of bone resected and extent of resection of soft tissue. Generally, it varies between 2.2% and 31.4% ⁷ , ⁹ , ¹¹ .

Influential factors

The first risk factor is tumor site. Due to the characteristics of the biological mechanics of the distal femur, the rate of aseptic loosening of prostheses in the distal femur is higher than for the proximal tibia. Wu et al. reported the rate of aseptic loosening of proximal tibia prostheses is 2.2% ⁷ , whereas Kawai et al. reported the rate of loosening for the distal femur is up to 21.9% ⁹ . The offset between the femoral axis and the force line crossing the centers of the femoral head and femoral condyle may act as an index for measuring bending moment. The offset is larger for proximal lesions, hence with these the bending moment is large (Fig. 8). In 1994, Duda et al. analyzed the mechanics of the human femur and reported the bending moment decreases from an average of 140 N•m at the level of the greater trochanter to 0 N•m at the point of insertion of the anterior cruciate ligament ²² . Stress on the interface between the prosthesis and host bone is distributed such that it is mainly concentrated on the tip of the prosthesis stem (Fig. 9). One study showed that 60% of the stress can transfer to the tip area of the prosthesis stem ⁸ . Therefore, the closer the distal femur prosthesis stem is to the proximal end, the greater the stress on the interface between the prosthesis and host bone, and the higher the incidence of prosthesis aseptic loosening.

Figure 8.

The offset between the line of force and the long axis of the femur decreases from the proximal to the distal femur.

Figure 9.

The stress distribution on the interface between prosthesis and host bone is such that stress mainly concentrates on the tip of the prosthesis stem (Black arrows).

A similar theory explains why the rate of proximal tibia prosthesis aseptic loosening is lower. Since tibia valga is only 3° in the standing position, the offset is very small. A transverse section of the tibial medullary cavity is close to triangular, unlike the orbicular‐ovate shape in that of the femur, thus the prosthesis stem fits the medullary cavity better.

The second risk factor is prosthesis design, which correlates closely with the incidence of aseptic loosening. Compared with hinged prostheses, rotary hinged prostheses permit 5°of intorsion or extorsion of the prosthesis stem at the tibial end; because this reduces the torsional force on the interface between the intramedullary prosthesis stem and host bone it can significantly reduce the incidence of aseptic loosening. Kawai et al. reported 82 cases of distal femur prosthesis reconstruction ⁹ . Among them, the rate of aseptic loosening in 51 cases with Lane‐Burstein hinged prostheses was 31.4%, whereas the rate in 31 cases with Finn rotary hinged prostheses was only 6.5%. Choong et al. reported only one case of aseptic loosening of the prosthesis in 32 patients with rotary hinged prosthesis reconstruction of knee joint with a 3.5 year follow‐up ²³ . What's more, hydroxyapatite or porous spray coating of the union of the prosthesis stem and host bone may also reduce aseptic loosening of prosthesis. Myers et al. analyzed the difference between rotary hinged and hinged prostheses in the prothesis revision rate due to aseptic loosening of the prosthesis ²⁴ . The prothesis revision rate was 35% in 162 cases with hinged prostheses, 24% in 15 cases with rotary hinged prostheses without hydroxyapatite spray coating and 0% in 158 cases with rotary hinged prostheses with hydroxyapatite spray coating. In accordance with the different anatomical features of these bones, the intramedullary stem of a femur prosthesis should have appropriate curvature, whereas the intramedullary stem of a tibia prothesis requires no curvature; thus the latter have a lower incidence of aseptic loosening of prostheses.

The last factor that influences the incidence of prosthesis aseptic loosening is the method of fixation of the prosthesis stem. The more stably the prosthesis is fixed, the smaller the stress in the local microenvironment and the lower the aseptic loosening rate. Now with the use of second or third generation bone cement the rate of prosthesis aseptic loosening is 2.2%–31.4%, whereas non‐bone cement fixation may significantly reduce the incidence rate of prosthesis aspect loosening. Griffin et al. reported 99 cases of prosthesis arthroplasty of knee joint fixed with non‐bone cement; the loosening rate of the prostheses was 2.0% ²⁵ . Flint et al. reported 44 cases of non‐bone cement fixation of proximal tibia artificial prostheses; over an average of 5 years follow‐up they reported no cases of loosening of prostheses ¹⁷ . However, Kawai et al. believe that the fixation form has no statistically significant effect on prosthesis survival rate ⁹ .

Diagnosis and treatment

Aseptic loosening of prostheses usually occurs within two years of surgery and manifests as progressively severe local pain; radiography shows a radiolucent line surrounding the prosthesis. The treatment methods include re‐immobilizing the prosthesis or replacing it, the choice is based on the condition of the patient's local bone and prosthesis. As for aseptic loosening of prostheses, the key point is prevention by taking appropriate measures in regard to immobilization forms, prosthesis design and so on to reduce the incidence of aseptic loosening.

Conclusion

Artificial prosthesis arthroplasty is the most common reconstruction approach after resection of malignant tumor surrounding knee joint. Although the rates of periprosthesis infection and aseptic loosening are high, they are acceptable. After artificial prosthesis reconstruction, the patient retains a complete limb and recovers satisfactory function. With the development of imaging techniques, maturation of surgical techniques, improvements in prosthesis design and application of combination adjuvant therapy, artificial prosthesis reconstruction will benefit more patients with malignant bone tumors around the knee joint.

Disclosure

No benefits in any form have been, or will be, received from a commercial party related to the subjects of this manuscript.