This retrospective clinical study was approved by the medical ethics board of our medical institute. Written or verbal informed consent was obtained from all patients. The indications for patellar fractures were as follows: congruity of the patellar surface > 2 mm, a gap between the patellar fragments > 3 mm, or inability to extend the knee due to tearing of the extensor retinaculum. Between January 2015 and February 2021, 94 displaced transverse (OTA/AO 34C1) patellar fractures were confirmed by radiography in our emergency room. All these fractures were surgically treated in our institute. Patients who met the below exclusion criteria were excluded from the current study: lost to follow-up before the 12-months post-operative follow-up (n = 2); impaired extension or flexion function of the ipsilateral knee joint before surgery (n = 2); unable to undergo rehabilitation at the medical instruction or unable to complete the final function evaluation due psychopathy or brain injury (n = 4); presence of an open patellar fracture (n = 1); previous fracture surgically treated on the ipsilateral lower extremity (n = 3); presence of a pathological patellar fracture (n = 0); aged less than 18 years (n = 1); treated with a technique other than the novel combined fixation technique described in this study (n = 47); presence of an old fracture (more than 14 days between the fracture and the surgery) (n = 3); did not provide written informed consent (n = 2); presence of a concomitant fracture or concomitant neurovascular injury on the ipsilateral lower extremity (n = 3). Thus, in total, 26 adult patients diagnosed with transverse patellar fractures treated with this combined fixation technique were included in this study. All surgeries were performed by one experienced orthopedic surgeon.
Each patient was placed in the supine position and a tourniquet was applied under anesthesia. An incision was made along the midline of the patella. After exposure of the fracture line, the hematoma in the fracture gap was removed. Then, several irrigations were performed to remove bone debris and fluid in the knee joint cavity. The anterior aponeurosis on the patella within 2 mm of the fracture line was elevated for exposure of the fracture line during reduction. Tight suture of the aponeurosis on the patellar surface was performed to facilitate and simplify the anatomic reduction. The reduction was checked to ensure it was satisfactory, and then temporary fixation was performed with a reduction clamp. After implantation of two longitudinal Kirscher wires, reaming in the smaller fragment was performed along the Kirscher wires in a monocortical fashion. Then, two cannulated screws were inserted. The length of each screw was shorter than the measurement to ensure that the tip of the screw would not protrude from the cortical bone or cut the suture. Then, cerclage and tension band wiring with non-absorbable polyester sutures (5 Ethibond-Ethicon Ltd., Edinburgh, UK) were performed in accordance with the Pyrford technique. During suturing, the knee joint was kept in extension. Every stitch was pulled forcefully and as close to the patella as possible. When suturing was completed, the stability of the fracture fixation was evaluated several times by full rotation of the knee joint; this was recorded in the surgical record and the physiotherapist was notified so as to inform individual rehabilitation. Before wound closure, the retinaculum was repaired, and the holes caused by reaming were sutured (see Figs. 1 and 2).
Figure 1. The combined fixation with cannulated screws and the modified Pyrford technique. (a) modified Pyrford technique; (b) cannulated screws fixation; (c) the combined fixation.
Figure 2. The image of transverse patellar fracture fixed with combined fixation during the surgery.
Post-operative rehabilitation was similar for all patients. Each patient had a cast applied with their leg in full extension for wound healing. Isometric quadriceps exercises and partial weight-bearing with crutches were encouraged immediately after surgery. The cast and stitches were removed simultaneously. Thereafter, active and passive knee flexion of 20 or 30 degrees, depending on the stability of the fixation, was instructed. If needed, the affected knee was supported by an adjustable knee brace. With tolerable pain, active knee flexion gradually increased 30 degrees every two to three weeks after progressive knee flexion training. At eight weeks post-surgery, the knee brace was removed, and the full range of knee motion was allowed. Importantly, knee rehabilitation was supervised by a physiotherapist to avoid secondary displacement of the fracture or failure of fixation. When bone union was confirmed by radiological examination, full weight-bearing without crutches was allowed.
The electronic medical records for all included patients were reviewed. The collected pre-operative data included body mass index (BMI), affected side, age, diagnosis of diabetes mellitus, gender, cigarette use, mechanism of injury, and the interval between injury and surgery. The included patients were followed up at one, two, three, six, and 12 months post-surgery, at a minimum. At each follow-up visit, plain radiographs were taken to assess the bone union time and post-operative complications. Further, the requirement for secondary surgery was recorded. The clinical evidence of patellar bone union included no tenderness with local palpation and the ability to continuously walk for three minutes without the aid of a crutch. The radiological evidence of patellar bone union included skeletal trabecula across the fracture line. Delayed bone union was defined as clinical and radiological evidence of bone union between three months and six months post-surgery. If the patient experienced delayed bone union, they were required to attend follow-up appointments every month until bone union was observed. The plain radiographs for all patients were assessed by two orthopedic surgeons. At the 12-months post-operative follow-up, the range of motion (ROM) of the affected knee joint was recorded and clinical outcomes were evaluated by the Bostman scoring system.