探讨活动型牛津膝单髁置换术在膝关节自发性骨坏死(SONK)治疗中的应用价值。

回顾性分析上海交通大学医学院附属松江医院骨科2015年1月至2022年8月收治的采用活动型牛津膝单髁置换术治疗的32例(32膝)SONK患者，其中单柱活动型牛津膝单髁置换14例，双柱活动型牛津膝单髁置换18例。术中患者麻醉成功后取仰卧位，屈膝90°取髌骨旁内侧至胫骨结节内侧入路，彻底清除股骨内髁病变坏死病变组织，根据病灶周围硬化骨完整度和强度选择截骨深度。小缺损以骨水泥填塞，较大缺损建议采用已清除的骨赘块制作细小碎骨柱填塞。安装单柱/双柱双柱活动型牛津膝假体。再次检查膝关节活动范围及稳定性，彻底止血，充分冲洗切口，关节周围注射鸡尾酒镇痛混合剂，常规放置引流管，逐层缝合，弹力绷带包扎固定。术后24 h内预防性应用抗生素，术后1~2 d拔除引流管，术后常规给予抗凝、镇痛等对症治疗，麻醉结束后主动进行股四头肌训练和踝泵练习。记录患者的手术时间及住院时间；记录患者术后1、3、6、12、24、36个月患膝并发症发生情况；分别于术前及末次随访时，收集并比较患者的膝关节功能情况[特种外科医院(HSS)评分、膝关节协会评分(KSS)、膝关节活动度]；测量并比较患者术前及末次随访时下肢力线情况[膝关节股胫角、胫骨内髁后倾角及胫骨平台角]；比较2种假体患者末次随访时的膝关节功能情况及下肢力线情况。数据比较采用t检验。

所有患者均顺利完成手术，手术切口均Ⅰ期愈合。32例患者的手术时间为35.0~70.0 min，平均手术时间(45.5±6.9) min；住院时间5.0~8.0 d，平均6.5 d。32例患者中，1例双柱活动型牛津膝单髁置换患者于术后1个月出现半月板垫片脱位，予以翻修后好转；1例行单柱活动型牛津膝单髁置换后，于术后24个月出现下肢力线改变和假体松动，翻修为全膝关节置换术后好转。其他患者均无血管神经损伤、围手术期感染、假体松动、半月板垫片脱位、下肢深静脉血栓形成等相关并发症发生。末次随访时患者膝关节HSS评分、KSS和膝关节活动度分别为(85.6±4.4)分、(88.4±5.2)分、(108.8±8.2)°，均显著高于术前[(66.8±5.2)分、(61.3±9.8)分、(97.5±7.6)°]，比较差异均有统计学意义(t＝15.61、13.81、5.71，P<0.05)；末次随访时患者膝关节股胫角、胫骨平台角分别为(174.4±5.6)°、(84.6±3.6)°，均显著低于术前[(179.5±3.8)°、(88.1±2.0)°]，比较差异均有统计学意义(t＝4.26、4.80，P<0.05)；膝关节胫骨内髁后倾角末次随访时为(82.8±3.4)°，较术前[(84.2±3.9)°]差异无统计学意义(t＝1.53，P＝0.131)。末次随访时，单柱活动型牛津膝单髁置换患者HSS评分、KSS、膝关节活动度、股胫角、胫骨内髁后倾角、胫骨平台角分别为(84.8±4.2)分、(89.2±6.0)分、(107.6±9.0)°、(175.8±6.0)°、(82.0±4.1)°、(83.8±3.2)°，与双柱活动型牛津膝单髁置换患者[(86.3±4.6)分、(87.8±4.5)分、(109.8±7.9)°、(173.4±5.4)°、(83.5±3.6)°、(85.3±4.0)°]比较，差异均无统计学意义(P>0.05)。

活动型牛津膝单髁置换在SONK患者的治疗中取得满意效果，并发症发生少，能明显改善患膝功能和部分纠正下肢力线，值得临床推广应用。

To explore the application value of movable Oxford unicondylar knee replacement in the treatment of spontaneous osteonecrosis of the knee (SONK).

Retrospective analysis was conducted on 32 patients (32 knees) with SONK admitted to the Department of Orthopedics, Songjiang Hospital Affiliated to Shanghai Jiaotong University School of Medicinel from January 2015 to August 2022, including 14 patients with single-column movable Oxford knee unicondyle, 18 cases of double-column movable Oxford knee unicondyle. During the operation, after successful anesthesia, the patient was placed in the supine position, bent the knee 90° to take the approach from the medial side of the patella to the medial side of the tibial tubercle, thoroughly removed the necrotic tissue of the medial femoral condyle, and selected the osteotomy depth according to the integrity and strength of the sclerotic bone around the lesion. Small defects should be filled with bone cement, and large defects should be filled with small bone fragments made from cleared osteophytes. The Oxford knee prosthesis with single/double column and double column movement was installed. The range of motion and stability of the knee joint were re-checked, hemostasis was thoroughly stopped, the incision was fully rinsed, the cocktail analgesic mixture was inject around the joint, the drainage tube was routinely placed, layer by layer suture, and elastic bandage was used for fixation. Antibiotics were used prophylactically within 24 h after surgery, drainage tube was removed 1-2 d after surgery, and symptomatic treatment such as anticoagulation and analgesia were routinely given after surgery. After anesthesia, quadriceps muscle training and ankle pump exercises were actively performed. The operation time and postoperative hospitalization time of patients were record; the incidence of knee complications was recorded 1, 3, 6, 12, 24 and 36 months after operation; the knee function [hospital for special surgery (HSS) score, knee society score (KSS) , knee range of motion] were collected and compared before surgery and at the last follow-up; the alignment of lower limbs [the femorotibial angle of knee joint, posterior tibial slope and tibial plateau angle] were measured and compared before surgery and at the last follow-up; the knee function and the alignment of lower limbs of patients with two prostheses were compared at the last follow-up. Data were compared by t test.

All the patients successfully completed the operation, and all the surgical incisions healed in stage Ⅰ. The operation time of 32 patients was 35.0-70.0 min, and the average operation time was (45.5 ± 6.9) min; the hospital stay ranged of patients was 5.0 -8.0 days, with an average of 6.5 days. Among the 32 patients, 1 patient with double-column movable oxford knee unicondylar replacement had a dislocation of meniscus pad one month after the operation, which was improved after revision; 1 patient underwent single-column movable oxford knee single condyle replacement, and the lower limb force line changes and prosthesis loosening occurred 24 months after the operation, the revision was improved after total knee replacement. Other patients had no serious complications such as vascular and nerve injury, perioperative infection, prosthesis loosening, meniscus pad dislocation, lower extremity deep vein thrombosis. At the last follow-up, HSS score, KSS and knee range of motion were (85.6±4.4) points, (88.4±5.2) points and (108.8±8.2)°, respectively, which were significantly higher than those before surgery [(66.8±5.2) points, (61.3±9.8) points, (97.5±7.6)°], the differences were statistically significant (t=15.61, 13.81, 5.71; P<0.05). At the last follow-up, the knee femorotibial angle and tibial plateau angle were (174.4±5.6)° and (84.6±3.6)°, respectively, which were significantly lower than those before surgery [(179.5±3.8)°, (88.1±2.0)°], the differences were statistically significant (t=4.26, 4.80; P<0.05); the posterior tibial slope of tibial medial condyle of knee was (82.8±3.4)° at the last follow-up, which showed no statistically significant difference compared with that before surgery [(84.2±3.9)°](t=1.53, P=0.131). At the last follow-up, the HSS score, KSS, knee range of motion, femorotibial angle, posterior tibial slope, and tibial plateau angle of single column movable Oxford knee unicondylar replacement patients were (84.8±4.2) points, (89.2±6.0) points, (107.6±9.0)°, (175.8±6.0)°, (82.0±4.1)°, (83.8±3.2)°, respectively, compared with thedouble column movable Oxford knee unicondylar replacement patients [(86.3±4.6) points, (87.8±4.5) points, (109.8±7.9)°, (173.4±5.4)°, (83.5±3.6)°, (85.3±4.0)°], there were no statistically significant differences (P>0.05).

The movable Oxford unicondylar replacement of knee has achieved satisfactory results in the treatment of patients with SONK, with fewer complications and can significantly improve the function of the affected knee and partially correct the alignment of lower limbs, which is worthy of clinical promotion and application.