Objective: The aim of the study was to analyze the clinical features of limb salvage surgery with epiphyseal preservation in children and adolescents, and to evaluate the recurrence rate, metastasis, complications, and the joint functional results after tumor resection. Methods: Between December 1995 and January 2003, 33 cases of preserving epiphysis procedure had been done. In this group, the tumor located in distal femur in 24 cases and in proximal tibia in 9 cases. There were 23 osteosarcomas, 6 Ewing’s sarcomas, 2 chondrosarcomas, and 2 aggressive osteoblastomas. The patients received 2–4 cycles of preoperative adjuvant chemotherapy, and another 6 cycles after surgery. The modified protocol T10 was taken. Radiography and MRI were used to determine the margins of the tumors preoperatively and histological examination was used to corroborate the evidence intraoperatively. The tumors were staged clinically. Among them, there were 2 cases in IA, 2 cases in IB, 17 cases in IIA, and 12 cases in IIB. According to the MRI analysis, the metaphyseal tumors in children were classified into 3 types. Type I, with the tumor close to but not contacting epiphyseal plate, and the distance between the two over 2 cm, was taken as absolute indication for the technique. Type II, with the tumor near or contacting epiphysis plate, and the distance between the two less than 1 cm, was taken as relative indication for the surgery. For type III, the tumor contacted the epiphyseal plate partially, and was over 2 cm beyond the joint. In the cohort, 18 cases were categorized as type I, 13 as type II, and 2 as type III. The size of the residual epiphyseal bone segment differed after different excision protocols which were taken according to the clinical classifications. Bone defects after tumor resections were repaired with massive intercalary allograft bone, followed by internal fixation by intramedullary nails and cancellous screws. Results: Among the 33 cases, 3 cases were lost to follow up. 29 cases had complete clinical data. Postoperative follow-up was 12–72 months. Recurrences were seen in 3 cases, with one local recurrence in type III cases one year post tumor resection, other 2 recurrences around the femoral vessels in type II cases 15 months and 30 months after the tumor resection respectively. The recurrence rate accounted for 10.34%. Amputation was performed for the cases with recurrences. But pulmonary metastasis developed and the patients died. In this cohort, 9 cases died. Five-year survival rate constituted 57.94% demonstrated by Kaplan-Meier sur- vival analysis. The median survival time for male patients was 64.36 months, and that for female patients was 56.00 months (P = 0.0403). The sites, stages and types were not associated with the survival time. Four cases reported 5 complications (17.24%). The complications included fracture of the allograft, limb length discrepancy, nerve injury, breakage of nails, and loosening of the screws. No patient reported skin necrosis, hematoma, infection, rejection to the allograft, and nonunion at the end of the follow-up. According to the functional evaluation criteria after surgical treatment of malignant tumors of the extremities introduced by Enneking, excellent functional results were reported in 11 cases, good in 13 cases, fair in 3 cases, and poor in 2 cases. The excellent or good result rate was 83%. The excision interface involved partial epiphysis in 4 cases, involved the epiphyseal plate in 8 cases. Intact epiphysis was preserved in 17 cases. Five cases with poor postoperative functional results were in types II and III. Average time of bone union was 2.5 months for epiphyseal end and 3.8 months for diaphyseal end. According to the International Society of Limb Salvage (ISOLS) radiological implants evaluation system, excellent reconstruction was observed in 23 cases, good reconstruction in 4 cases, fair reconstruction in 1 case, and poor reconstruction in 1 case. All patients in the group had satisfactory joint stability. No patient reported joint dislocation, valgus or varus deformity, and osteoarthritis. The average limb length discrepancy was 3.2 cm, ranging 2–6 cm. Conclusion: The limb salvage surgery with preservation of epiphysis for malignant bone tumors in children and adolescents guarantees the patient’s satisfactory postoperative limb functional results. Preoperative effective adjuvant chemotherapy and prevention of postoperative complications deserve great attention. Objective: The aim of the study was to analyze the clinical features of limb salvage surgery with epiphyseal preservation in children and adolescents, and to evaluate the recurrence rate, metastasis, complications, and the joint functional results after tumor resection. Methods: Between December 1995 In this group, the tumor located in distal femur in 24 cases and in proximal tibia in 9 cases. There were 23 osteosarcomas, 6 Ewing’s sarcomas, 2 chondrosarcomas, and 2 aggressive The patients received 2-4 cycles of preoperative adjuvant chemotherapy, and another 6 cycles after surgery. The modified protocol T10 was taken. Radiography and MRI were used to determine the margins of the tumors preoperatively and histological examination was used to corroborate the evidence The patients were staged clinically. Among them, there were 2 cases in IA, 2 cases in IB, 17 cases in IIA, and 12 cases in IIB. According to the MRI analysis, the metaphyseal tumors in children were classified into 3 types. Type I, with the tumor close to but not contacting epiphyseal plate, and the distance between the two over 2 cm, was taken as absolute Type II, with the tumor near or contacting epiphysis plate, and the distance between the two less than 1 cm, was taken as relative indication for the surgery. For type III, the tumor contacted the epiphyseal plate partially, and was over 2 cm beyond the joint. In the cohort, 18 cases were categorized as type I, 13 as type II, and 2 as type III. The size of the residual epiphyseal bone segment differed after different excision protocols which were taken according to the Clinical classifications. Bone defects after tumor resections were repaired with massive intercalary allograft bone, followed by internal fixation by intramedullary nails and cancellous screws. Results: Among the 33 cases, 3 cases were lost to foll ow up. 29 cases had complete clinical data. Postoperative follow-up was 12-72 months. Recurrences were seen in 3 cases, with one local recurrence in type III cases one year post tumor resection, other 2 recurrences around the femoral vessels in type II Cases of 15 months and 30 months after the tumor resection respectively. The recurrence rate accounted for 10.34%. Amputation was performed for the cases with recurrences. But this pulmonary disease has developed and the patients died. In this cohort, 9 cases died. Five-year survival Rates were 57.94% demonstrated by Kaplan-Meier sur-vival analysis. The median survival time for male patients was 64.36 months, and that for female patients was 56.00 months (P = 0.0403). The sites, stages and types were not associated with the Four complications reported 5 complications (17.24%). The complications included fracture of the allograft, limb length discrepancy, nerve injury, breakage of nails, and loosening of the screws. No patient reported skin necrosis, hematoma, infection, rejection to the allograft, and nonunion at the end of the follow-up. According to the functional evaluation criteria after surgical treatment of malignant tumors of the extremities introduced by Enneking, excellent functional results were reported in 11 cases , good in 13 cases, fair in 3 cases, and poor in 2 cases. The excellent or good result rate was 83%. The excision interface involved partial epiphysis in 4 cases, involving the epiphyseal plate in 8 cases. Intact epiphysis was preserved in Seven cases. Five cases with poor postoperative functional results were in types II and III. Average time of bone union was 2.5 months for epiphyseal end and 3.8 months for diaphyseal end. According to the International Society of Limb Salvage (ISOLS) radiological implants evaluation system , excellent reconstruction was observed in 23 cases, good reconstruction in 4 cases, fair reconstruction in 1 case, and poor reconstruction in 1 case. All patients in the group hNo patient reported joint dislocation, valgus or varus deformity, and osteoarthritis. The average limb length discrepancy was 3.2 cm, ranging 2-6 cm. Conclusion: The limb salvage surgery with preservation of epiphysis for malignant bone tumors in children and adolescents guarantees the patient’s satisfactory postoperative limb functional results. Preoperative effective adjuvant chemotherapy and prevention of postoperative complications deserve great attention.