With the establishment of surgical staging, the development of reconstruction methods and the gradual improvement of effective chemotherapy, the treatment of malignant bone tumors of the extremities has made great progress in recent years, and the treatment has evolved from the amputation-based approach in the past to the current surgical treatment based on limb preservation therapy (1).From July 1997 to July 2005, we treated 167 patients with osteosarcoma of the lower extremities with limb preservation therapy, and all patients were treated with preoperative After preoperative standardized chemotherapy, all patients underwent extensive resection and tumor-based prosthetic reconstruction, and 100 of them were followed up. This paper reports the following information on the follow-up of patients, the summary of surgical results and related complications. 1. Clinical data 1.1 General information There were 56 male cases and 44 female cases in this group. The age ranged from 13 to 57 years. Location: proximal femur in 5 cases, distal femur in 57 cases, and proximal tibia in 38 cases. The routine preoperative examinations included X-ray plain film of the affected limb, CT and MRI, CT of the lung, and whole-body bone scan. Among them, 3 cases had Ennecking stage IIA, 85 cases had IIB and 12 cases had III. All patients underwent neoadjuvant chemotherapy and postoperative chemotherapy, two cycles of chemotherapy before surgery and three to five cycles of chemotherapy after surgery depending on the necrosis rate of the tumor specimen. The main chemotherapeutic agents were adriamycin, cisplatin, high-dose methotrexate and isocyclophosphamide, including adriamycin 60-80 mg/m2, cisplatin 100-120 mg/m2, methotrexate 8-12 g/m2, and isocyclophosphamide 12.5 g/m2, of which cisplatin was administered by arterial cannula. Cisplatin was administered by arterial cannula, and the other drugs were administered intravenously. 1.2 Surgical method Preoperatively, the anatomical data were measured according to X-ray plain film and the artificial prosthesis was customized. According to CT and MRI, the extent of lesion involvement was determined, and the osteotomy plane was 3-5 cm from the edge of the tumor, and the lateral approach was mostly used for tumor resection and artificial joint reconstruction in the upper femur, with longitudinal splitting of the broad fascial tensor muscle to reveal the greater trochanter, separation and resection along the normal tissues around the swelling, and after joint dislocation, the osteotomy plane was judged preoperatively to cut off the backbone and complete resection of the tumor, followed by prosthesis implantation. After the joint dislocation, the tumor was completely resected by cutting the backbone according to the preoperative osteotomy plane, and then the prosthesis was implanted and the gluteus medius muscle reconstruction was routinely performed. Lower femoral tumor resection prosthesis reconstruction is often performed by using a medial curved knee incision, entering along the gap between the medial femoral muscle and the suture muscle, resecting the tumor anteriorly within the normal rectus femoris muscle, separating along the middle femoral tumor outside the normal tissue, paying attention to protecting the femoral artery vein, and resecting the tumor together with the surrounding normal muscle; cutting the medial and lateral collateral ligaments, cruciate ligaments and meniscus, dislocating the knee joint; pulling the femur, cutting the femur The femur was pulled, the muscles attached to the posterior aspect of the femur were cut, the femoral stem was separated upward, and the stem was also cut according to the preoperative osteotomy plane, and the lower femoral prosthesis and tibial plateau were implanted. The upper tibial preservation surgery is often performed by using a medial knee incision, entering along the inferior border of the medial femoral muscle to the femur, cutting the joint capsule, cutting the medial periosteum of the tibial tuberosity, maintaining the continuity of the patellar ligament and the fascia below, turning the patella laterally along with the patellar tendon, cutting the medial and lateral collateral ligaments, cruciate ligament and meniscus, and dislocating the knee joint. The tumor of the upper tibial segment was separated, the tumor and surrounding normal tissues were removed together, the tibia was truncated according to the preoperative scope, the femoral condyles were preserved and the lateral femoral prosthesis was installed, and the upper tibial segment prosthesis was installed and reset. For patients who needed partial bone segment inactivation and reimplantation, the tumor was extensively resected, and the bone segment to be reimplanted was sawed off and all the tumor tissue on the tumor bone was removed, then treated with 20% hypertonic saline for 20 min, 95% alcohol for 20 min, and inactivated for a total of 40 min. 17 of the 100 patients in this group used this method in combination with artificial prosthesis to reconstruct the bone defect, and 21 patients used allograft bone to reconstruct the bone defect. The remaining 62 patients were reconstructed with artificial prostheses. In this group, 71 cases of domestic prosthesis (Chunli Zhengda Company) and 29 cases of imported prosthesis (LINK Company, Germany) were used. The MSTS 93 (Musculoskeletal Tumor Society 93) limb preservation scoring system was used to evaluate the function of the follow-up patients after surgery.(2) The scoring system included 6 items: pain, overall function, acceptance, support, walking function, and gait, with a total of 30 points out of 5 points for each item. The current retrospective study scored 24-30 for excellent function, 18-23 for good function, 12-17 for moderate function, and less than 12 for poor function. Percentages are commonly used to express the functional score results (patient score divided by 30 points). All patients were followed up for 1-8 years, with a median follow-up time of 3.5 years. The 3-year survival rate was 81.8% and the 5-year survival rate was 65%. 6 patients had prosthesis fracture from 4 months to 12 months after surgery (2 with imported prosthesis and 4 with domestic prosthesis); 13 patients had delayed prosthesis infection (12 with domestic prosthesis and 1 with imported prosthesis); 2 patients had prosthesis loosening; 5 patients had graft-host bone junction not Local recurrence occurred in 7 patients within 0.5 to 2 years after surgery, 4 of which were soft tissue tumor recurrence and underwent tumor re-excision, after which the tumor was satisfactorily controlled and no recurrence occurred; the other 3 patients had amputation. No fatigue fracture of the graft bone occurred in this group. All patients started to move the joint actively and practice quadriceps muscle 1 week after surgery; 2 weeks later, they started to walk on the ground with the help of crutches, and the joint function was good; six months later, the functional reconstruction score was above 23(2), and the overall score of 100 patients at follow-up was 23.30±5.17 (77.66±17.22%), and the patients’ limb function was excellent 62%, good 27%, moderate 7%, poor 4%, and the excellent rate was The excellent rate was 89%. The development of neoadjuvant chemotherapy has created conditions for limb-preserving treatment, thus preoperative chemotherapy is the precondition for limb-preserving surgery. Pre-operative chemotherapy can result in obvious tumor demarcation, reduced vascularity and increased necrosis; clinically, tumor shrinkage, reduced fever and increased movement of adjacent joints can be seen; increased calcification, tumor volume reduction and surrounding edema can be seen on imaging (3,4). Before 1980, less than 20% of patients with malignant bone tumors underwent limb-preserving surgery. The main purpose of limb-sparing surgery is to provide proper local control of the primary tumor, but limb-sparing resection must ensure that the local recurrence rate is no higher than that of amputation and that it produces a good functional outcome. Among the methods of reconstruction after resection of malignant bone tumors of the limb, tumor-based prosthetic replacement has the best functional outcome (5,6). 3.1 Functional evaluation of tumor-based artificial joint replacement Postoperative function of limb-preserving surgery is determined by various aspects such as tumor size, extent of resection, surgical technique of reconstruction, prosthesis design, rehabilitation exercise effect, patient’s subjective motivation and degree of cooperation. For the hip joint, the reconstruction of abduction function is very important for postoperative function, and there is no significant difference between the effect of bimanual hip reconstruction and total hip reconstruction on postoperative function (7). For the proximal femur composite prosthesis has better postoperative function than the prosthesis alone (8). Age, pathological fracture and prosthesis site have been noted in the literature as important factors affecting postoperative function after periprosthetic tumor preservation (9). Our study utilized the MSTS 93 functional reconstruction scoring system [4] to rate limb function in 100 patients with osteosarcoma who were followed up to date. Two of the scores, pain and support, had a perfect score of 79% and 70%, respectively, meaning that the patients did not have any postoperative pain symptoms and did not need any support for walking. The majority of patients scored 3 to 4 on the remaining four scores. The highest functional score was 24.67±1.15 for the upper femur, followed by 23.65±4.98 for the lower femur, and the lowest functional score was 22.97±5.6 for the upper tibia. 67% to 83% of MSTS 93 scores were reported in the literature for patients with lower extremity tumors after limb preservation using tumor-based artificial joints (10). Our follow-up results showed the best postoperative function in patients with osteosarcoma of the upper femur. Compared with other reconstruction methods, tumor-based prosthetic reconstruction had more satisfactory postoperative function than previous reconstruction methods. 3.2 Survival analysis of tumor-based lower limb prosthesis The 3-year survival rate of tumor-based prosthesis in this group was 81.8%, and the 5-year survival rate was 65%. The Kaplan-Meier survival curves for this group of prostheses are shown in Figure 4. Factors affecting prosthesis survival include tumor site, percentage of osteotomy, gender, age, prosthesis stem diameter and reconstruction method. The 5-year survival rate for the upper femoral prosthesis has been reported in the literature to be 63% to 90% [11]. The intact rate of composite prosthetic reconstruction at this site was significantly better than that of simple prosthetic reconstruction, with a 10-year survival rate of 77% to 84% for composite prosthesis (11), compared with 65% for simple prosthetic reconstruction. No complications with the prosthesis occurred in the five patients of the upper femoral segment in our follow-up, and the prognosis was better. The 5-year intact rate of tumor-based prosthesis in the distal femur has been reported in the literature as 66%-83% (12), and the 3-year and 5-year survival rates of tumor-based prosthesis in the 57 patients with distal femur followed up by this group were 80.3% and 69.1%, respectively. The 5-year survival rate of the upper tibial prosthesis was generally 54% to 60% (15), which is the worst artificial joint survival rate among the three common sites of tumors in the lower extremity, probably due to the poor stability of the prosthesis compared with the upper and lower femoral segments on the one hand, and the high incidence of prosthetic complications due to the difficulty of soft tissue coverage on the other hand. The 3-year and 5-year survival rates of the upper tibial prosthesis in our group were 79% and 57.1%, respectively. Our findings suggest that the prognosis of tumor-based prostheses differs in different parts of the lower extremity, with the best prognosis for the upper femur, followed by the lower femur, and the worst prognosis for the upper tibial prosthesis. 3.3 Complications of tumor-based prosthesis 3.3.1 Delayed infection of the prosthesis This complication mostly occurs about 1 year after surgery. Patients repeatedly experience redness, swelling, heat and pain at the surgical site, which is aggravated by walking and relieved by rest, and finally sinus tracts appear, and the fluid flowing out is cloudy, but the bacterial culture is mostly negative. Among the 13 cases of delayed infection of prosthesis in this group, 12 cases were domestic prosthesis and 1 case was imported prosthesis, indicating that there are still quality problems of domestic tumor prosthesis, probably mainly due to the processing technology. The main cause of delayed infection is that the artificial prosthesis irritates the surrounding soft tissues and produces a chronic inflammatory reaction, which gradually forms sinus tracts. In 13 cases of revision of prosthesis due to delayed infection, most of the soft tissues around the prosthesis were black in color, with a large amount of inflammatory granulation tissue growth. In four cases, a one-stage prosthesis revision was performed, and after removal of the prosthesis, the inflammatory granulation tissue around the prosthesis was thoroughly removed, the cement in the medullary cavity was removed, repeatedly flushed, soaked with dilute iodine, and the prosthesis was reinserted after autoclaving. Two of them were successful, and no re-infection occurred at 3 years of follow-up; one case was replaced with a new prosthesis in one stage, and re-infection occurred 1 year later; the other case re-infected 1.5 years after surgery. In 6 patients, the prosthesis was removed and temporarily replaced with bone cement, and a new prosthesis was reinserted 1 year later, and 5 of them were successful. The remaining 3 patients underwent amputation. 3.3.2 Fracture of the prosthesis stem In two cases of the knee prosthesis from Link, Germany, the fracture occurred at the joint between the stem and the femoral condyle, which was a problem in the design of the prosthesis, and the design has been changed. In two other cases, the fracture occurred at the junction between the endosteal portion of the stem and the extraosseous segment of the prosthesis (root of the stem). This type of fracture is related to the quality of the prosthetic material and the thickness of the stem. In general, the diameter of the root of the stem should not be <12 mm around the knee joint (13). 2 cases of fracture occurred in the shaft of the joint, which is related to the design of the prosthesis, and the fractured prostheses were all distal femoral prostheses, which may be related to the extensive resection of the distal femur, the inability to preserve the lateral collateral ligament of the knee joint, and the excessive stress on the prosthesis. 3.3.3 Aseptic loosening of the prosthetic stem Aseptic loosening of the prosthesis is the most common complication of tumor-based prosthesis replacement (14, 15), which is characterized by progressively increasing pain in the knee and radiographic findings of osteolysis and translucent areas around the prosthetic stem. Aseptic loosening of the prosthesis usually occurs 4 to 5 years after surgery. In this group of patients, because the follow-up period is still short, prosthetic loosening does not occur much. There are many factors that influence the loosening of the prosthesis: (1) the location of implantation, the lowest loosening rate is in the proximal femur, followed by the proximal humerus, distal femur and finally the proximal tibia (16); (2) the age of the patient, the failure rate is significantly higher in young patients; (3) the amount of bone tissue removed, the loosening rate is significantly higher in those with more than 40% of the distal femur removed than in those with less than 40% (17). In this group, composite artificial joints were used for those who had more than 50% of bone tissue removed, so the effect of the amount of non-removed bone tissue on the loosening of the prosthesis was not shown (18). The loosening rate of cement-fixed prostheses was lower than that of cement-based prostheses (19, 20). Prosthetic sagging occurs mostly in cases where prosthetic loosening has already occurred. This may be mainly due to the gravitational conduction concentrated at the junction between the prosthesis and the level of the osteotomy, causing the bone at the junction to gradually dissolve resulting in sinking of the prosthesis. Fractures after artificial prosthesis replacement often occur at the tip of the stem of the prosthesis, which is related to the malposition of the stem implant, straight stem prosthesis, etc. Fractures often occur in the femur because of the forward physiological curvature of the femur. The average MSTS 93 functional reconstruction score was above 23 in our group of 100 patients at six months, and the tumor-based knee joint preserved the best knee function compared with other methods of limb-preserving reconstruction, such as hemiarticular inactivation and reimplantation. The local recurrence rate was less than 7%, indicating that the surgical method was reasonable, but the complication rate was still high, and the design and processing of the tumor-based prosthetic joint needed further improvement.