Proximal humeral fractures are relatively common in the clinical work of orthopedics. According to the literature, they account for 4% to 5% of all fracture cases, and the morbidity increases year by year. More than 71% of proximal humeral fracture cases involve the patient of 60 years or above. Neer pointed out that most of the proximal humeral fractures were slightly displaced or had no obvious displacement. With conservative treatment, the fractures above could achieve bone healing and satisfactory function. However, according to the latest multicenter study, more than 64% of proximal humeral fractures were displaced. Among all displaced proximal humeral fractures, approximately 40% were Neer type 2-part and were recommended for surgical treatment. For elderly patients or patients with severe underlying disease who cannot tolerate surgery, the conservative treatment can be taken. The advantages of traditional surgical approach include wide surgical field and anatomical reduction, but the functional recovery is affected by wide range of local soft tissue dissection, large blood supply damage and postoperative adhesion. Domestic and foreign literature studies have shown that the minimally invasive treatment of proximal humeral fractures was effective, and the proximal humerus internal locking osteosynthesis system (PHILOS) has unique advantages.

I. General information: 1. Inclusive criteria: (1) Closed fracture, the greater tuberosity fracture (displacement of >5 mm) or surgical neck fracture (angulation of >45° or displacement of >1 cm) in X-ray films without neurovascular injury; (2) Treatment of MIPPO technique combined with PHILOS internal fixation; (3) Visit within 2 weeks after injury and 12 months or more of follow-up. 2. Exclusive criteria: (1) Combination of other injuries; (2) Surgery intolerance due to severe internal medicine disease; (3) Pathological fractures. 3. Imaging evaluation: standard X-ray films, including scapular anteroposterior view, scapular lateral view and west point view. For the fractures with unclear X-ray film, thin-layer CT scan of 2 mm and 3D reconstruction are feasible. MRI can provide a more reliable interpretation of shoulder surrounding soft tissue. Especially for elderly patients with a history of shoulder dislocation, MRI is necessary to determine whether there is rotator cuff injury. From July 2014 to March 2016, 20 patients including 12 males (60%) and 8 females (40 %) with proximal humeral fractures were treated with minimally invasive percutaneous plate osteosynthesis (MIPPO) technique combined with PHILOS internal fixation, and the average age was (45.95±11.40) (22-63) years. 10 cases were 2-part greater tuberosity humerus fracture, which included 5 males (50%) and 5 females (50%) with an average age of (42.10±13.79) (22-62) years. Among them, 1 case had distal radial fracture. 10 cases were 2-part surgical neck humerus fractures, which included 7 males (70%) and 3 females (30%) with an average age of (49.80±8.13) (36-63) years. Among them: 1 case had distal radial fracture; 1 case had ulnar olecranon fracture; 1 case had multiple rib fracture. Causes of injury: 15 cases of fall damage and 5 cases of traffic accident. The time from injury to surgery ranged from 5 to 10 days with an average of 7.5 days. The X-ray film indicated greater tuberosity fractures or surgical neck fractures of humerus with displacement. This study was approved by the hospital ethics committee and all subjects signed informed consent. II. Surgical treatment: 1. Anesthesia: general anesthesia combined with brachial plexus block. 2. Surgical position: beach chair supine position with the affected shoulder joint extending beyond the outer edge of operating table and the cushion under scapula. 3. Surgical procedure: a longitudinal incision of 4 cm was made under the anterior angle of acromion. The skin, subcutaneous tissue and deep fascia were cut open layer by layer, and the approach was selected between the anterior bundle and the middle bundle of deltoid muscle. As the axillary nerve and its course were touched with finger, a suture was made with 4# silk thread at the distal end of incision for nerve protection. The wire retractors were carefully used to avoid nerve damage caused by excessive traction. The clavipectoral fascia was cut open to expose proximal humerus. No. 5 Ethbond suture was used to sew the rotator cuff tendon attached to greater and lesser tuberosities, and the greater tuberosity fractures or surgical neck fractures of humerus were reduced and then fixed temporarily with 1 or 2 Kirschner wires of 2.0 mm. A 5-hole AO 3.5 mm PHILOS plate of appropriate length was selected, and a longitudinal incision was made at the corresponding position of the anterolateral site of upper arm. The skin, subcutaneous tissue and deep fascia were cut open to expose proximal humerus through the space between deltoid and biceps. The upper and lower tunnels were perforated, and the plate was inserted from the proximal end to the distal end while staying close to periosteum. As the plate was confirmed at the center of shaft, a total of 7 or 8 screws was used for fixation in the distal and proximal ends. Under C-arm fluoroscopy, the anterior (internal, external and central position) and axillary views were conducted to check the positions and lengths of the screws, the stability of fracture ends during passive anteflexion and upward lifting, abduction and the internal and external rotation of shoulder joint, the acromial impingement syndrome and the arthralgia. The rotator cuff was repaired if damaged. Combined fractures of other parts in ipsilateral extremity should be operated simultaneously. After operation, the deltoid muscular tension was examined to determine whether the axillary nerve was damaged. III. Postoperative management: The incision was closed layer by layer, and both patients were required to place a drainage tube. The drainage tube was removed on the 1st postoperative day according to the volume of drainage, Antibiotics were used prophylactically for 1 day in both groups. The affected limb in patients with rotator cuff injury was immobilized with shoulder abduction and external rotation brace for protection of 6 weeks, and the active exercises of wrist, elbow and fingers of the affected limb were allowed on the 1st day after surgery. The affected limb in patients without rotator cuff tear was immobilized with forearm sling for protection of 4 weeks, and the pendulum exercises and passive anteflexion and upward lifting of shoulder joints were allowed on the 1st postoperative day. All patients started limb-assisted shoulder joint activities in the 2nd week, and active shoulder joint exercises in the 3rd week. The shoulder abduction and upward lifting were controlled within 60° 3 weeks after operation, and 90° after 4 weeks. As X-ray films showed osteophyte growth 6 weeks after operation, the sling was removed to perform active shoulder joint exercises. At 3 months after surgery, the affected limb could perform daily activities without limitation. IV. Efficacy evaluation indicators: All surgery was performed by the same group of doctors, and the outpatient visit and evaluation were conducted by another two senior orthopedic surgeons. The patients were followed up closely after discharge from the hospital, and regularly reviewed at 1st, 2nd, 4th, 6th, 8th and 12th months after operation. The anteroposterior view, lateral view and west point view X-ray films of the shoulder joint were taken according to the situation. The operation time, blood loss, internal fixator position, fracture healing and functional status of the two groups was compared. The shoulder joint function was assessed using shoulder joint mobility, rating scale of the American shoulder and elbow surgeons (ASES) , and Constant-Murley score. The pain of affected shoulder joint was evaluated using visual analogue scale (VAS) . Among them, 0 point was divided into completely painless, 1-3 points was divided into mild pain, 4-6 points was divided into moderate pain and 7-10 points was divided into severe pain. The ASES score is the shoulder joint function evaluation criteria developed by ASES, including pain (50%) and life function (50%) , with a total score of 100 points. The higher the score, the better the shoulder function. The Constant-Murley score is used for shoulder joint evaluation, including pain (15 points) , daily life (20 points) , deltoid muscle strength (25 points) , and range of motion (40 points) ; 90-100 points are considered excellent, 80-89 points are considered good, 70-79 points are considered moderate and <70 points are considered poor. V. Statistical software and methods: SPSS22.0 software was used for data analysis. The quantitative data were described by the mean±standard deviation, and the independent sample t test was used for statistical inference. The qualitative data were described by percentage, and the Chi-square test was used for statistical inference; A value of P≤0.05 was regarded as statistical difference.

Twenty patients were followed up for 12 to 24 months with an average of (15.50±9.66) months. The mean follow-up time of patients with greater tuberosity fractures was (16.05±7.92) months. The mean follow-up of patients with humeral surgical neck fractures was (14.40±10.88) month. All patients attained fracture healing without the complication of axillary nerve injury, humeral head necrosis loosening plate, broken nail, delayed union, nonunion or infection. Among them, 12 were male and 8 were female (P=0.515) . The mean ages of the two groups were (42.10±13.79) (36-63) years and (49.80±8.13) (22-62) years (P=0.146) respectively; the mean operation times were (120.50±27.43) (90-160) mins and (133.90±46.41) (60-200) mins (P=0.442) respectively; the mean fracture healing times were (2.25±0.54) (1.5-3.0) months and (2.60±0.81) (1.5-4.0) months (P=0.270) respectively; the mean VAS scores were (0.40±0.52) (0-1) points and (0.50±0.53) (0-1) points (P=0.673) respectively; the mean ASES scores were (94.64±3.31) (90-98.3) points and (91.65±5.76) (85.0-98.3) points (P=0.172) respectively; the mean Constant-Murley scores were (95.10±3.12) (91-99) points and (92.60±5.62) (83.5-99.0) points (P=0.235) respectively; the mean postoperative anteflexion and upward lifting angles were (174.00°±5.16°) (170°-180°) and (167.00°±9.49°) (150°-180°) (P=0.055) respectively; the mean external rotation angles were (43.00°±8.23°) (30°-50°) and (34.00°±10.75°) (20°-50°) (P=0.050) respectively; the mean abduction angles were (158.00°±13.98°) (130°-180°) and (149.00°±19.69°) (110°-170°) (P=0.254) respectively.

The treatment of Neer 2-part proximal humeral fractures with MIPPO technique combined with PHILOS internal fixation was clinically validated. There was significant difference in postoperative external rotation angle restoration between the cases of 2-part humerus surgical neck fractures and 2-part humerus greater tuberosity fractures. However, there was no significant difference in operation time, gender, age and postoperative recovery.