比较肱骨近端粉碎性骨折应用锁定钢板联合同种异体腓骨植骨与半肩关节置换术的疗效。

回顾性分析海军军医大学附属长海医院2011年5月至2014年12月收治的老年复杂肱骨近端骨折患者（Neer分型：3型、4型）共60例，其中锁定钢板联合同种异体腓骨植骨患者28例（A组），半肩关节置换患者32例（B组）。随访期间测量肱骨头高度、颈干角。患肩评价采用Constant肩关节评分标准及视觉模拟评分（visual analogue scale，VAS），记录患者肩关节活动度及术后并发症。

术后随访时间13~18个月，A组平均（15.5±1.8）个月，B组平均（15.2±2.2）个月；末次随访A组在Constant肩关节评分、肩关节的活动范围明显高于B组，术后并发症的发生数少；术后肩关节疼痛VAS评分，B组更有优势；术后影像学评估，B组肱骨头高度丢失及颈干角变化与A组无明显差异；根据Paavolainen方法，A组优良率大于B组。

同种异体腓骨植骨重建肱骨近端骨折内侧柱,术中联合肱骨近端锁定钢板能有效支撑肱骨头，预防肱骨头塌陷及螺钉穿出,较半肩关节置换具有更好的临床疗效及更低的并发症。在解决疼痛方面,肩关节置换治疗效果更好。

Proximal humeral fracture is one of the most common fractures. According to statistics, the incidence of proximal humeral fractures accounts for about 4% - 5% of all fractures and 45% of humeral fractures, most of which are older patients with osteoporosis. If the reduction cannot be as effective as possible, especially for the old proximal humeral fracture of three or four parts, it may produce joint stiffness, humeral head necrosis and other complications, which directly affecting shoulder function . At present, there are many treatment options for this type of fracture, but the treatment results are still controversial . Compared with traditional internal fixation, locking plate is considered to be a very effective choice for the treatment of osteoporotic proximal humeral fractures , but for patients with medial unstable fractures, locking plate alone cannot achieve good fixation. During the follow-up, complications such as humeral head collapse, screw cut out and humeral head necrosis were found, which seriously affected shoulder function. However, for osteoporotic patients with four-part fracture, humeral head splitting, anatomical neck fracture, shoulder dislocation, old fracture more than 6 months, and some serious three-part fracture, shoulder replacement can also be selected. More and more reports have proved that shoulder replacement has a satisfactory effect on the recovery of shoulder function after operation. In recent years, the use of locking plate-assisted fibular allograft to reconstruct the medial column of proximal humeral fracture has achieved good results, and biomechanics has also proved that it has high mechanical stability . The purpose of this study is to compare the effect of proximal humeral comminuted fractures treated with shoulder replacement and allogenic fibular graft combined with locking plate through retrospective analysis, so as to provide the basis for future surgical choice.

Ⅰ. General information. From May 2011 to December 2014, 60 patients with comminuted proximal humeral fracture were treated in our hospital. They were divided into two groups according to the operation methods: 28 patients were treated with locking plate combined with allogenic fibula graft, and 32 patients were treated with hemiarthroplasty. 1. Inclusion criteria: (1) unilateral closed proximal humeral fractures; (2) Neer classification of three-part and four-part fractures; (3) unstable or comminuted medial calcar fractures; (4) age 62-84 years. 2. Exclusion criteria: (1) open fracture; (2) pathological fracture; (3) old fracture; (4) Neer classification of two-part fracture; (5) past shoulder joint dysfunction; (6) past history of rotator cuff injury. Of the 60 patients, 29 were males and 31 were females, ranging from 62 to 84 years old, with an average age of (72.3±6.6) years. Causes of injury: 30 cases of fall, 15 cases of traffic accidents, 9 cases of sports injuries, 6 cases of falling injuries. The time from injury to operation was 3-6 days, with an average of (3.4±1.3) days. Before operation, all patients underwent AP and lateral radiographs of the affected shoulder joint and three-dimensional CT reconstruction to determine the type of fracture. Zimmer Biomet proximal humerus locking plate was used as the proximal humerus plate and the prosthesis was Zimmer Biomet cement modular standard humeral head. Ⅱ. method. (1) Locking plate combined with allogenic fibula graft group: After general anesthesia or brachial plexus anesthesia, patients took beach chair position with shoulder pads. After routinely disinfect by iodophor and set up, the deltoid-pectoral approach was used. Skin and subcutaneous tissue were incised layer by layer. During the operation, attention was paid to protecting the cephalic vein. After blunt separation along the deltoid and pectoralis major space, we exposed the proximal fragments and cleared the hematoma. Intertubercular grove was used as an anatomical marker to tract the upper limb, pry and reposition the humeral head, and Kirschner wire was temporarily fixed. Depending on the medial unstable fracture and the diameter of the bone marrow cavity, the appropriate allogenic fibula was inserted into the distal bone marrow cavity, and then the fracture was reduced. During the operation, the allogenic fibula was placed on the medial side of the humeral head and the Kirschner wire was temporarily fixed. Under fluoroscopy, the proximal humerus locking plate can be placed on the lower edge of the greater tuberosity for about 5 mm, and the lateral to the intertubercular sulcus for about 5-10 mm. A lag screw was inserted in advance of the sliding hole to determine the appropriate height of the plate and then the distal and proximal locking screw is placed in turn. The drilling depth should be accurately measured. The tip of screw should be about 5 mm below the articular cartilage. Finally, the distal end of the plate was fixed with bicortical locking screw and the Kirschner wire was removed. After examining and repairing the muscles and rotator cuff attached to the tuberosities, the negative pressure drainage was placed. The incision was closed routinely. The forearm shoulder strap was used after operation. (2) Hemiarthroplasty group: Anesthesia mode, position and surgical approach were the same as locking plate group. After exposure of the fracture, blood clot was fully cleared to identify the greater and lesser tuberosities. For the patients with avulsion fracture of great and less tuberosities, sutures should be used to mark them and rotator cuff. The rotator cuff should not be separated from the greater and lesser tuberosities. With the humeral head as the reference, the appropriate size of prosthesis was selected. The medullary cavity was gradually reamed. The appropriate humeral head and humeral stem trial were installed. Attention should be paid to adjusting the humeral neck shaft angle and the humeral head retroversion to 20-40 degrees to check the stability and range of motion of the shoulder joint. After washing the medullary cavity, bone cement was implanted and the artificial humeral head prosthesis of the size of the trial was installed. Reduction of greater and lesser tuberosities and surrounding fracture fragments, with absorbable sutures to fix them to the proximal end of the humeral shaft, and rotator cuff injury was carefully repaired. The drainage tube was placed and the incision was closed layer by layer. The forearm shoulder strap was used after operation. (3) Postoperative management:Routine antibiotics were used for one day after operation. Pain control was routinely used. Patients were encouraged to start passive exercises of shoulder joint 2 weeks after operation; X-ray examination was conducted 6 weeks after operation to evaluate the recovery, and active and resistance exercises were performed; and shoulder weight-bearing exercises began 12-13 weeks after surgery. (4) Observations and Functional Evaluation:The operation time and intraoperative bleeding volume of the two groups were counted; the height of humerus and neck-shaft angle were measured by imaging; the shoulder function was scored by Constant shoulder score and visual analogue score (VAS) . The range of motion of shoulder was also recorded during the last follow-up. The complications such as failure of internal fixation, screw cut-out, necrosis and infection were recorded during the follow-up period. (5) Statistical analysis: SPSS 21.0 statistical software was used to analyze the statistical data. T-test was used for measurement data and was expressed by (±s) . The utilization rate of counting data (%) was expressed by χ² test, and the difference was statistically significant (P<0.05) .

28 patients in locking plate group and 32 patients in hemiarthroplasty group were followed up for 13 to 18 months, with an average of (15.3±2.3) months. Among them, the average follow-up time of locking plate group was (15.5±1.8) months, and hemiarthroplasty group was (15.2±2.2) months. There was no significant difference in follow-up time between the two groups (t=0.5781, P=0.5655) . The age of locking plate group was (73.2±6.4) years old, and hemiarthroplasty group was (72.5±6.9) years old. There was no significant difference between the two groups (t=0.4074, P=0.6852) . There was no statistical difference in gender composition, injury mechanism and time from injury to operation between the two groups. In terms of postoperative complications, 1 case of humeral head necrosis and 1 case of internal fixation protrusion occurred in locking plate group during the follow-up period; 2 cases of skin infection, 2 cases of greater tuberosity displacement and 3 cases of joint stiffness occurred in hemiarthroplasty group. In terms of postoperative complications, locking plate group had fewer complications than hemiarthroplasty group.Sixty patients were followed up after operation, and the functional data of the two groups were recorded and evaluated at the last follow-up. The range of flexion, external rotation and internal rotation of the affected shoulder in locking plate group were 146°±18° (91°-167°) , 50°±13° (30°-70°) , 49°±7° (24°-71°) ; and 120°±20° (43°-82°) , 40°±15° (33°-66°) , 42°±11° (24°-70°) in hemiarthroplasty group respectively. The data of three groups were statistically significant (P<0.05) . There was significant difference in shoulder activity between two groups.Constant shoulder score was 74.5-83.0 in locking plate group at the last follow-up, with an average of (74.4±5.1) ; Constant shoulder score was 60.3-76.5 in the hemiarthroplasty group, with an average of (64.8±4.0) , P<0.05. There was a significant difference in Constant score between the two groups. At the last follow-up, VAS score of locking plate group assessed shoulder pain symptoms by (4.0±0.7) and VAS score of hemiarthroplasty group assessed shoulder pain symptoms by (2.0±1.0) , P<0.05, with significant difference.The imaging results showed that the average loss of humeral head height in locking plate group was (2.0±0.6) mm at the last follow-up, and that in hemiarthroplasty group was (1.9±0.3 ) mm, P=0.409, with no significant difference between the two groups; the neck-shaft angle in locking plate group at the last follow-up was 128.1°±10.5° (89°-141°) ; and the neck-shaft angle in hemiarthroplasty group was (130.4°±4.2°) , P=0.259, with no significant difference between the two groups. According to Paavolainen method (the neck-shaft angle is excellent at 130°±10° and good at 100°-120° and poor at less than 100. At the last follow-up, 19 cases were excellent, 6 cases were good and 3 cases were poor, the excellent and good rate was 89.2%; 13 cases were excellent, 9 cases were good and 10 cases were poor in hemiarthroplasty group, the excellent and good rate was 68.8%. The joint function evaluation in locking plate group was better than that in hemiarthroplasty group.

Although the use of allogenic fibula transplantation increases the cost of treatment, we have a plenty of fibula models to choose to adapt to different patients with medullary cavity and fracture, improve the efficiency of surgery, shorten the operation time, reduce the risk of surgical infection. What's more, it can effectively support the humeral head, prevent the collapse of the humeral head and screw penetration. The shoulder function of the patients recovered good after the operation, greatly reducing the complications after the operation. It is superior to hemiarthroplasty in function and has satisfactory short-term clinical effect.