Humeral shaft fracture refers to the rupture between the surgical neck of 1-2 cm below and the humeral condyle of 2 cm above, which mostly occurs in the middle of the humeral shaft, followed by the lower part and the upper part. The lower third humeral shaft fracture is clinically common and associated with radial nerve injury. It is often caused by indirect violence and presented as oblique or spiral, which is prone to nonunion. Due to the high incidence of radial nerve injury and nonunion, the treatment is more intractable. Alternative methods for surgical procedure involve external fixator, intramedullary nail, plate, etc., and internal fixation and open reduction is generally considered to be the most reliable treatment method. The standard surgical approaches for open reduction and plate fixation include anterior lateral approach, anterior approach, anterior MIPPO technique approach, posterior approach, and medial approach (ulnar approach) . The literature reported that the incidences of iatrogenic radial nerve injury caused by the anterolateral approach, the anterior approach, and the posterior approach were 1/5, 1/25 and 1/9, respectively. In the meanwhile, according to other articles, the incidence of nonunion of this fracture type after open reduction was as high as 15%. The causes of nonunion are attributed to traumatic factors, surgical dissection, etc., but this is mainly due to the fact that single plate fixation does not provide sufficient stability for some "special" fractures. For this purpose, this paper puts forward the specialty of ''lower third humeral shaft fracture'', emphasizing the necessity of double plates for such fractures. In the meanwhile, the ulnar approach and the anterior MIPPO approach with dual minimally invasive incisions and double plates for rigid internal fixation of lower third humeral shaft fracture were proposed in view of the high incidence of iatrogenic radial nerve injury. The active and passive elbow extension and flexion started without any brace for immobilization. The results were satisfactory without iatrogenic radial nerve or ulnar nerve injuries, and all fractures healed as scheduled.

I. Clinical research: (1) General information: A total of 20 patients (14 males and 6 females) with an average age of 34.4 (19-60 ) years were selected for this research. There were 6 left cases and 14 right cases. The causes of injury included 6 cases of automobile accidents, 6 cases of hand wrestling injuries and 8 cases of fall damages. All patients were closed injury, among which 1 case was combined with femoral fracture, 1 case was combined with contralateral radial fracture, and 1 case was combined with pelvic fracture. According to the AO/ASIF classification, there were 3 cases of type A, 12 cases of type B and 5 cases of type C. (2) Surgical methods. After general anesthesia, the patient was placed in the supine position with routine disinfection and draping. The affected upper arm was abducted in 90°, and the forearm was in supination and 80° of flexion, which was placed on the operating table. A longitudinal incision was made on the inside of the upper arm, and the skin, subcutaneous tissue, and fascia were cut open layer by layer to expose the space between musculus biceps brachii and musculus triceps brachii. The ulnar nerve was exposed and separated carefully, and the rubber sheet was used to pull the ulnar nerve medially for protection. The median nerve and brachial artery were pulled anteriorly for protection. The fracture ends were debrided, reduced and fixed temporarily with Kirschner wires after exposure. A 3.5 mm reconstruction plate or anatomic locking plate was used for fixation of the medial distal humerus with 3 screws on the proximal side and 2 screws on the distal side. As fracture was well reduced under fluoroscopy, a longitudinal incision was made at the coracoid of the proximal anterior upper arm and along the pectoralis major deltoid sulcus. After undermining dissection of the distal humerus, a 3.5 mm LCP of 10-12 holes was inserted with Kirschner wires for temporarily distal and proximal fixation. As the position of the plate was well placed under fluoroscopy, the musculus biceps brachii, musculus brachialis, brachial artery, and median nerve was pulled laterally under direct vision with Hoffman wire retractor to expose approximately 3/5 of the anteromedial area of the distal humerus. The distal plate was placed slightly on the medial side with MIPPO technique (for easy screwing) with 2 locking screws on the distal side for fixation. The proximal screw was inserted under direct vision, and the other two screws were placed percutaneously. The fracture ends were confirmed well reduced under fluoroscopy without elbow joint entrapment or ulnar coronal process impingement. After irrigation, the incision was closed layer by layer. The sheet rubber was placed medially, and the wound was covered with the sterile dressing. Approximately 6 hours after the operation, the patient was allowed fisting exercise. On the 1st postoperative day, the patient was able to perform active and passive elbow extension and flexion, mainly the active exercises, which included shoulder abduction, upthrow and pendulum exercises. II. Observation and follow-up indicators: The operation time, intraoperative blood loss, ulnar nerve function of the radial nerve, and fracture healing time were observed postoperatively. Neer shoulder function score was used to evaluate shoulder joint function, and Mayo elbow function score was used to evaluate elbow joint function.

The operation time of 20 patients in this group was 60-110 min with an average of (85.0±6.5) min. The blood loss was 30-60 ml with an average of (53.0±7.2) ml. All the incision healed in the 1st stage without iatrogenic radial nerve injury, ulnar nerve injury, medial nerve injury or brachial artery injury. One patient had symptoms of radial nerve injury before surgery without postoperative progression after surgery, and the signs recovered 3 months later. The fracture healing time was 10-15 weeks with an average of (13.0±2.1) weeks. The maximum elbow flexion range was 131°-146° (137.60°±3.51°) , and the maximum elbow extension range was 0°-5° (2.70°±0.91°) . The Mayo elbow function score was 80-100 (92.04±5.72) points, and the Neer shoulder function score was 85-100 (93.63 ± 4.11) points.

The specialty of the lower third humeral shaft fracture should be emphasized. It is recommended to perform double plate internal fixation to reduce the incidence of delayed union and nonunion. The ulnar approach combined with anterior MIPPO double incision was proposed in view of the higher incidence of iatrogenic radial nerve injury in the distal humeral incision. Our study shows that the ulnar approach combined with anterior MIPPO approach has the advantages of minimal invasion, rigid fixation and avoidance of iatrogenic radial nerve injury in the treatment of the lower third humeral shaft fracture.