This retrospective cohort study was conducted at a single academic center between March 2017 and January 2021. All patients who underwent anterior ankle arthroscopy to assess possible syndesmotic instability were screened ( Fig 1 ). Inclusion criteria were arthroscopically confirmed syndesmotic widening ≥2 mm, persistent symptoms for ≥3 months postinjury, documented preoperative patient-reported outcome measures (PROMs), and at least 1 year of postoperative follow-up.

Flow diagram of patient selection. Initial arthroscopic assessment for syndesmotic instability was performed in 86 patients. After applying exclusion criteria and confirming syndesmotic widening ≥2 mm with persistent symptoms for ≥3 months, 41 patients (43 ankles) completed postoperative patient-reported outcome measures (PROMs) at median follow-up of 2.3 years. A subset of 31 patients (33 ankles) completed additional functional testing and radiographic assessment.

Exclusion criteria included concomitant ankle fracture, previous syndesmotic or osteochondral surgery, more than 2 previous operations on the affected ankle, significant preoperative nerve injury, and pre-existing ankle arthritis. The Regional Ethics Committee approved the study (approval no. [264009]), and all participants signed an informed consent.

Patients were generally referred because of severe, persistent ankle pain and markedly reduced function after trauma. Often, patients could not recall a clear injury event until questioned in detail. When described, injury mechanisms included not only the classical external-rotation pattern but also supination or axial-load trauma, and sometimes remained unclear. On prompting, many patients recalled acute signs such as a distinct “pop,” significant early swelling, bruising (laterally, medially, or proximally), and prolonged inability to bear weight. Many had initial radiography to exclude a fracture. Chronic symptoms typically involved persistent swelling, push-off weakness, stiffness, instability (giving-way), and pain that could be lateral, medial, anterior, or diffuse.

For this study, we extracted clinical signs explicitly documented in patient records: tenderness over the anterior or posterior tibiofibular ligaments, and a positive Cotton test, defined as palpable side-to-side laxity or ≥2 mm lateral talar translation observed during mortise fluoroscopy.

Preoperative imaging including unilateral non−weight-bearing radiographs, bilateral weight-bearing radiographs, CT scans, or MRI was performed at the surgeon’s discretion. Diagnostic conclusions were determined on the basis of finalized musculoskeletal radiology reports from various radiologists and institutions; original images were not re-evaluated.

Contemporary standards for syndesmotic assessment cite specific thresholds for abnormality. For instance, radiographic evaluations, with measurements taken 10 mm above the tibial plafond may identify a tibiofibular clear space >6 mm (on the anteroposterior view), tibiofibular overlap <6 mm (anteroposterior view) or <1 mm (mortise view), or medial clear space >4 mm as abnormal findings. On bilateral weight-bearing radiographs, a side-to-side difference >1 mm in these parameters was also considered indicative of instability. Typical CT criteria define a side-to-side difference >2 mm in the tibiofibular interval as abnormal, with measurements taken from axial slices 10 mm above the tibial plafond. MRI may identify ligament discontinuity, abnormal morphology (such as waviness or thickening), or increased signal intensity within the syndesmotic ligaments. ^,,

However, as reports originated from multiple radiologists and institutions, the precise diagnostic criteria applied in each case were not investigated. This was underscored by observed inconsistencies; for example, some CT reports identified side-to-side differences of 1 to 2 mm as pathological, indicating some heterogeneous practice. Because isolated clinical tests and routine imaging may be inconclusive, or even normal, once an injury becomes chronic, arthroscopy was offered when three criteria were met: (1) a history suggestive of syndesmotic trauma, (2) persisting major symptoms (pain with markedly limited function) for more than three months, and (3) no alternative diagnosis that better explained the symptoms. Immediately before arthroscopy, fluoroscopic stress views estimated talar tilt (0–5°, >5-15°, or >15°) and any syndesmotic or medial clear space widening (>2 mm).

All operations were performed with the patient supine under total intravenous anesthesia without routine muscle relaxation. A thigh tourniquet, applied preoperatively before draping, was typically inflated only after the diagnostic arthroscopy portion and immediately before suture-button placement.

Anterior ankle arthroscopy was performed without joint distraction, using a standard 4.0-mm 30° arthroscope introduced through an anteromedial viewing portal. Instruments were passed via an anterolateral working portal. Fluid pump pressure was maintained at 25 to 35 mm Hg. The tibiofibular interval was probed dynamically with a 4-mm trocar, a 4-mm shaver, or a 2- to 4-mm hook probe to establish the minimal central gap, in a manner consistent with principles described by Guyton et al. The arthroscope itself was occasionally leveraged to tilt the talus carefully to aid visualization of widening, avoiding iatrogenic cartilage damage. No manual external-rotation stress was applied to the foot during this assessment. Syndesmotic widening of ≥2 mm was considered pathologic. Ankles in which posterior widening was observed to clearly exceed the central widening were classified as having a posterior-predominant pattern.

When widening met the ≥2 mm threshold, scar tissue within the syndesmotic interval was excised. To restart the healing process in ligaments that had scarred in an elongated state, the anteroinferior tibiofibular ligament was typically divided centrally with scissors. The posteroinferior tibiofibular ligament was partially released superiorly from its tibial attachment with a 4.0 shaver (Bone Cutter 4.0; Arthrex, Naples, FL), preserving its distal-most fibers to guide healing in an anatomically reduced position. In some ankles with minimal talar tilt, accessing the full extent of the posterior syndesmosis for release could be challenging.

Stabilization was achieved with either 2 identical suture-buttons or a hybrid combination (TightRope, Arthrex; INVISIKNOT Non-Fracture, Smith & Nephew, Andover, MA). These were inserted through separate 3 to 5 cm longitudinal lateral and medial incisions. The first tunnel was drilled from the fibula, just proximal to the syndesmosis, aiming slightly more proximal on the tibia; the second tunnel was placed approximately 1 to 2 cm above the first. Care was taken during lateral drilling to preserve an anterior fibular bony bridge. The medial incision was guided by the marked exit of the drill tip and facilitated direct visualization of the tibial exit points, allowing for protection of the saphenous vein and nerve from impingement. Correct and firm button seating against the cortex were confirmed by fluoroscopy, by direct visualization and palpation. Three reinforcing knots were typically tied over the most proximal button after solid sequential tightening, even if the suture-button was “knotless.” Arthroscopic reassessment of syndesmotic reduction after final tightening was not routinely performed, but compression of the syndesmosis was crudely assessed on fluoroscopy comparing side to side before and after suture button compression.

A modified Broström procedure with suture-tape augmentation was added if intraoperative stress fluoroscopy showed talar tilt visually estimated to be 10 to 15° or greater. This combined procedure also was performed if lateral ankle instability had been the primary preoperative diagnosis but syndesmotic widening (≥2 mm) was identified arthroscopically, even if the visually estimated talar tilt appeared to be less than 10 to 15°.

We used no casting or bracing postoperatively. Patients bore limited weight (10-20 kg) for 3 weeks before progressing to full weight-bearing as tolerated. Follow-up visits were scheduled at 6 weeks and 3 months. Suture-button removal was considered if patients reported discomfort, generally no earlier than one year postoperatively.

The primary outcome was the change in Foot Function Index (FFI) and Foot and Ankle Ability Measure (FAAM) scores from preoperative levels to at least 1 year postoperatively. Preoperative PROMs were obtained by the first author during the diagnostic process, completed together with the patients, and documented in the electronic patient record. The FFI assesses pain and functional limitations using a 17-item scale scored from 0 to 10, whereas the FAAM evaluates limitations in activities of daily living (21 items) and sports activities (7 items) using Likert scales (0-4) converted to 0 to 100. Each measure also includes a global score, an FFI limitation score (0-10) and FAAM-ADL and FAAM-Sport function scores (0-100), yielding a total of 6 PROM subscores.

At least 1 year postoperatively, patients completed follow-up PROM assessments via telephone interviews conducted by an independent assessor (J.J.) not involved in the surgical care. During these interviews, patient satisfaction was also recorded on a 0 to 10 scale categorized as very dissatisfied (0-2), dissatisfied (3-4), neutral (5-6), satisfied (7-8), or very satisfied (9-10). All follow-up PROM and satisfaction data were documented in the electronic patient record.

Secondary outcomes included suture-button removal rates and complications, obtained from the electronic patient journal. A subset of patients returned for functional and radiographic assessment conducted by independent clinicians (J.J., M.A.M.) not involved in the surgery ( Fig 1 ). These evaluations occurred shortly after the PROM follow-up assessments and included single-leg stance balance, heel raise endurance, and ankle dorsiflexion range of motion. Bilateral weight-bearing ankle radiographs were also obtained at this time to assess tibiotalar and medial clear space, and anterior tibial bony spur formation.

Pre- and postoperative PROMs, along with demographic, injury, surgical, and complication data were retrieved from the electronic patient journal (EPJ; DIPS, Bodø, Norway). Imaging records were accessed via the Picture Archiving and Communication System (Sectra, Linköping, Sweden), and fluoroscopic stress test results were documented in the EPJ. Body mass index, American Society of Anesthesiologists status, and surgical time were recorded in the surgical planning program Orbit (Evry, Oslo, Norway). Double data entry ensured accuracy of PROM and satisfaction scores.

Sensitivity of imaging modalities was calculated by dividing the number of positive findings for syndesmotic pathology on each modality by the total number of ankles examined with that modality, using arthroscopic confirmation of ≥2 mm widening as the reference standard.

Statistical analyses were performed using R version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria) and Python 3.12 (Python Software Foundation), with P <.05 considered significant. Before applying paired t tests, normality of data distributions was assessed using the Kolmogorov-Smirnov test. If distributions were non-normal, Mann-Whitney U tests were used. Missing data were minimal; no imputation was performed, and analyses were determined using available data only. No formal power analysis was conducted as this was a retrospective study. The minimal clinically important difference (MCID) was defined as 75% of preoperative score standard deviation, rather than the conventional 50%, due to the lack of established MCID values for the 6 PROMs in chronic syndesmotic instability. This more stringent threshold increases confidence in the clinical significance of observed improvements.