Inclusion criteria: (1) patients who were diagnosed with chronic lateral ankle instability (CLAI) and received standard conservative treatment for over 6 months, with no improved symptom; (2) patients with unilateral ankle ATFL injury (non-avulsed fracture type) but no calcaneofibular ligament injury (confirmed under arthroscopy); (3) The fibular side of ATFL remnant was integral, as verified under arthroscope, and the ATFL could not be repaired directly; (4) patients who were willing or required to perform all arthroscopic ATFL reconstruction, and the transplanted tendon was the autologous semitendinosus tendon; (5) patients who received ankle ligament surgery for the first time, with no previous history of ankle injury or surgery; (6) patients with complete surgical and follow-up data, and were followed up for at least 24 months clinically.
Exclusion criteria: (1) patients combined with foot and ankle deformity, abnormal line of force, fracture, ankylosis and other ligament injuries; (2) those combined with central and peripheral neuromuscular atrophic disease or ligament laxity; (3) patients with ankle osteoarthritis or osteochondral injury that required osteochondral transplantation; (4) those with severe underlying disease and could not tolerate the surgery.
From January 2015 to July 2017, a total of 53 CLAI patients undergoing ATFL reconstruction were enrolled into the current study according to the above inclusion and exclusion criteria. The institutional review boards of our hospital approved our study protocol. All cases provided the informed consent as well as consents for the Health Insurance Portability and Accountabilty Act to participate in this study. All operations were completed by the same surgeon.
The above patients were randomized as remnant-preserving group (preservation of the ATFL remnant, n=25) and non-preserving group (no preservation of the ATFL remnant, n=28) according to the random number table method. There was no statistical difference in the general preoperative data between two groups; in addition, differences in Visual Analogue Scale (VAS) score, the American Orthopedic Foot and Ankle Society (AOFAS) score, Karlsson Ankle Functional Score (KAFS) , and Anterior Talar Translation (ATT)  between two groups were not statistically significant. (Table 1)
The patient was in supine position, and a 7-cm cushion was placed in the affected hip after successful anesthesia, so that the affected leg was in the neutral position of ankle under natural anesthesia status. Then, the affected leg was placed in the distal edge of the operating table for convenient operation technique, the pressure pneumatic tourniquet was put in the middle segment of thigh, and the pneumatic pressure was set at 60 kPa after blood evacuation.
First of all, the ipsilateral semitendinosus tendon of patient about 6-8 cm in length was excised, the soft tissue and muscular tissue on the tendon were completely cleaned, and only the tendinous tissue was preserved. Subsequently, the tendon was subjected to 10 min of pre-load at 150 pounds. A button plate (Smith & Nephew, USA) that was 15 mm in length was selected, the tendon was penetrated into the button plate at equivalent length, and the both sides of tendon were then sutured.
The ankle anterior-lateral and anterior-medial standard approaches were adopted to explore and treat the lesion within the ankle joint, and the scars as well as inflammatory hypertrophic tissues were also cleared. Afterwards, the anterior subtalar approach was constructed to explore the ATFL, and the Footprint region on the fibula was also exposed. For the remnant-preserving group, the remnant on the ATFL fibula was preserved (Fig 1a); for the non-remnant-preserving group, the ATFL tendinous tissue was completely cleaned. Then, a guide pin was inserted into the Footprint region center of fibular from the anterior-inferior to the posterior-superior direction through this channel, which formed a 30°angle with the fibular longitudinal axis (Fig 1b). Later, the 4.5 mm hollow drill was used to drill a bone tunnel, which should be long enough to penetrate the fibula (Fig 1c/1d). Then, the Footprint region on the talus was explored and cleaned through the ankle anterior-lateral approach and the anterior subtalar approach, a guide pin was inserted through the anterior subtalar approach, and the 6.0 mm hollow drill was utilized for drilling to form a bone tunnel that was 20 mm in depth.
The tendon graft was put into the lateral ankle through the anterior-lateral ankle approach (Fig 2a), later, the guide pin with a wire was used to penetrate the button plate into the fibular tunnel (Fig 2b), and then the other terminal of tendon graft was pulled back to better confirm the button plate fixing effect (Fig 2c). Afterwards, under the neutral position of ankle joint, the grafted tendon was strained, and a 6.0 mm interface screw (Smith & Nephew, USA) was used to fix the distal end of grafted tendon inside the talar tunnel (Fig 2d and Fig 3).
The short-legged bracket was used to fix the ankle joint for 2 weeks at mild dorsiflexion and eversion position. On the second day after surgery, the patient was advised to conduct early non-weight-bearing functional exercises or isometric exercises of lower limb muscles. Two weeks later, weight-bearing functional exercises were carried out in the presence of ankle AircastTM (DJO, Vista, CA, USA), and physical activities were encouraged after removing the Aircast at 6 weeks later.
Postoperative follow-up and observational indexes
The patient wound healing and ankle stability were followed up after surgery; besides, the VAS, AOFAS, KAFS and ATT scores were adopted to assess the ankle function. The ankle proprioception was assessed using the Active Joint Position Sense (AJPS) . All measurements were completed by the same rehabilitation physician who did not participate in surgery and was blind to grouping results.
The SPSS 19.0 statistical software was used for analysis. The measurement data, such as VAS, AOFAS, KAFS, ATT, and AJPS scores, in each group before and after surgery, as well as between two groups after surgery, were compared and analyzed using t test (symmetric distribution) or Mann-Whitney test (asymmetric distribution). Pearson chi-square test was adopted to compare the enumeration data, like age and sex. A difference of P<0.05 was deemed as statistically significant.