Introduction

Ankle sprains are one of the most common injuries among the general population and during the practice of sports. Forced ankle inversion is the most common mechanism of injury. The anterior talofibular ligament (ATFL), followed by the calcaneofibular ligament (CFL), are the most commonly injured ligaments. Although patients with acute ankle sprains can be successfully managed conservatively, approximately 15–20% of patients remain symptomatic and eventually develop chronic ankle instability (CAI) and experience repetitive ankle sprains [4, 13, 34].

Patients with CAI may have both medial and lateral ankle symptoms [19]. Moreover, deltoid ligament abnormalities have been described in patients with CAI [2, 10, 19, 21]. It is estimated that 40% of patients with CAI have a partial deltoid injury [19, 30]. The concept of rotational ankle instability (RAI) involves a combination of lesions in the medial and lateral ligamentous complexes; this has therapeutic and prognostic implications [5]. Under weightbearing conditions, ATFL-deficient ankles show a significant increase in anterior translation, internal rotation, and superior translation of the talus [6]. These biomechanical changes are associated with the development of degenerative arthritis. CAI may also cause concomitant ankle injuries, leading to a deltoid ligament tear. RAI should be considered when an injury is present in the anterior fibers of the deltoid ligament. RAI is often difficult to assess clinically, particularly if patients lack medial-sided ankle symptoms. Should these patients be treated with an isolated repair of the lateral ligaments, medial symptoms may then present and become apparent. It is, therefore, always advisable to check for a deltoid ligament injury and have this treated accordingly in the same procedure, alongside the lateral collateral ligament repair or reconstruction.

Recently, some authors have developed arthroscopic techniques to treat CAI with excellent results [1, 9, 17, 23, 26, 32, 33, 35]. However, no such techniques have been published to treat injuries of both lateral and medial ligaments in patients with RAI.

The aim of the study was to describe an arthroscopic all-inside repair of the lateral collateral ankle ligament complex and deltoid ligament injuries with a knotless suture anchor technique. The initial results in a series of 13 patients were presented. The described technique represents the first combined lateral and medial ankle ligament repair procedure, both performed entirely arthroscopically.

It was hypothesized that an “open book” anterior deltoid ligament tear would be a common feature in patients with symptoms of RAI, and that an arthroscopic repair would yield excellent results.

Materials and methods

Between 2013 and 2015 (a 3-year period), 125 patients were arthroscopically treated for ankle instability by the senior author. From this group, 13 patients were diagnosed with RAI [12 males and 1 female, median age 32 years (range 15–54 years)] during arthroscopy and were surgically treated during the same procedure. RAI is defined as an abnormal increase of talar rotation within the tibiofibular mortise, due to an injury in the most anterior component of the deltoid ligament secondary to a chronic deficiency of the lateral collateral ligament. Diagnostic criteria included evidence of ligamentous injury of both the lateral and medial ligamentous complexes confirmed arthroscopically. The right ankle was affected in 8 of the 13 cases.

All patients included in the study had sustained more than one previous ankle inversion injury, and had completed a minimum of 6 months of protocolized physiotherapy. All patients reported pain or discomfort in the lateral ankle, as well as a subjective sensation of ankle instability on exertion. Only ten patients reported pain (6) or discomfort (4) in the medial ankle. Two patients reported posterior ankle pain. None of the patients had sustained any foot or ankle fractures, or had undergone any previous foot or ankle surgery. Physical examination showed no foot deformity. Lateral ankle instability was reproduced during clinical examination (with a positive anterior drawer test); however, no signs of medial instability were elicited in any of the cases. No patients reported any discomfort around the peroneal, posterior tibial, or Achilles tendons. Functional outcomes using the American Orthopedic Foot and Ankle Society (AOFAS) hindfoot score were assessed preoperatively and at the latest follow-up (minimum of 1 year after the procedure).

Plain radiographs showed an os trigonum in one of the cases and a tibial osteophyte in another patient. Preoperative magnetic resonance imaging (MRI) of the ankle was obtained in all cases. MRI showed chronic ATFL injury in all cases. A calcaneofibular ligament injury was detected in one patient, and a chronic tear of the deltoid ligament in seven patients. Osteochondral injury at the talar dome was observed in two cases, one medial and one lateral. One patient had loose bodies in the posteromedial aspect of the ankle.

The study was approved by the ethical committee of the Hospital Quirón Barcelona (number: 10-03-2016).

Operative technique

Instruments used for this technique included a 4.0 mm 30° arthroscope, an arthroscopic 3.5 mm motorized shaver and burr, and standard arthroscopic instruments. Suture passers (Microsuture lasso curved 70°, Arthrex, Naples, FL, USA, and MiniScorpion, Arthrex, Naples, FL, USA), 0 non-absorbable sutures (Fiberwire, Arthrex, Naples, FL, USA), and knotless suture anchors (Pushlock 2.9 mm × 15 mm, Arthrex, Naples, FL, USA) were used for ligamentous repair. Introduction of a cannula (PassPort Button cannula, 6 mm ID × 2 cm, Arthrex, Naples, FL, USA) into the anterolateral and anteromedial portals helped provide an easy passage of instruments and sutures, and reduced the likelihood of injury to at-risk anatomical structures (i.e., the superficial peroneal nerve in the anterolateral portal).

Under spinal anaesthesia, the patient was positioned supine with the affected leg on a thigh support placed under the knee.

An ankle dorsiflexion arthroscopic technique without distraction was performed, where anteromedial and anterolateral ankle portals were established.

Where there was significant synovitis or scar tissue, synovectomy or lysis of adhesions with shaver was carried out. Both gutters were then examined, and the lateral collateral and deltoid ligaments probed to assess the anatomical characteristics.

The lateral collateral ligament was examined with the arthroscope introduced through the anteromedial portal, and the probe inserted through the anterolateral portal. The ATFL upper band is observed on the floor of the lateral recess of the ankle. In dorsiflexion, the ATFL is relaxed and can be observed as a hammock from the fibula to the talar neck. This anatomic feature is critical in the recognition of an ATFL injury. A lax ATFL due to a mid-substance ligament injury is observed as a wider-than-normal lateral gutter. Where all or part of the ATFL fibular footprint is observed, a complete or partial detachment of the ATFL is indicated, respectively.

The deltoid ligament was examined with the arthroscope introduced through the anterolateral portal, and the probe through the anteromedial portal. The structures of the medial gutter—the anterior aspect of the medial malleolus, the distal tip of the medial malleolus, the medial wall of the talus, the joint capsule, and the deltoid ligament—must be carefully observed and recognized. The medial malleolus on its anterior view usually consists of a thin fringe of cartilage that articulates with the medial wall of the talus, followed by a central fringe of cortical bone, and the insertion of the anterior and deep components of the deltoid ligament. The tip of the medial malleolus is free of ligamentous insertion. On the floor of the medial recess, the deep layer of the deltoid ligament is observed. In dorsiflexion, the deep component of the deltoid ligament is observed as a hammock that originates on the medial wall of the talus and inserts onto the medial and distal aspects of the medial malleolus. These anatomic features are crucial for recognition of a deltoid ligament injury (Fig. 1).

Fig. 1
figure 1

Anatomical perspective of the ankle. Arthroscopic views of lateral and medial gutters showed to depict normal appearance (A—lateral gutter and, C—medial gutter), and pathological appearance with a ligament rupture (B—ATFL detachment and, D—deltoid ligament “open book” injury). (1) ATFL. (2) Fibular footprint of the ATFL (highlighted with black dotted line). (3) Anterior area of the deltoid ligament. (4) Deltoid ligament “open book” detachment

Full size image

After full arthroscopic examination, the deltoid ligament was repaired. Subsequently, the lateral collateral ligament was repaired using an all-inside technique as previously reported [35]. Any other necessary arthroscopic procedures were performed before the repair of both ligaments.

The deltoid ligament was repaired arthroscopically through the anterolateral portal, directing the arthroscope towards the medial gutter. The footprint for the attachment of the deltoid ligament on the medial malleolus had to be debrided with a shaver introduced through the anteromedial portal (Fig. 2). A suture passer was introduced through the anteromedial portal. Under direct arthroscopic visualization, the area of the ligament desinserted was penetrated from medial to lateral. Although this step could have been performed with a non-automatic suture passer (Microsuture lasso curved 70°, Arthrex, Naples, FL, USA), the authors’ preferred technique utilised an automatic suture passer clamp (MiniScorpion, Arthrex, Naples, FL, USA). By folding the suture thread in half (double suture), a loop and two ends were obtained. The double suture was placed first into the automatic clamp. It was then introduced with the automatic suture passer through the anteromedial portal and the ligament was pierced by the suture. Once the clamp was removed, the loop and the suture ends were in the same (anteromedial) portal. Both suture ends were introduced into the loop; by pulling the ends, the loop entered the joint and was locked, while the ligaments were grasped by the suture (Fig. 3).

Fig. 2
figure 2

Arthroscopic view of the medial gutter in a right ankle. Scope introduced through the anterolateral portal and directed to the medial gutter. Shaver introduced through the anteromedial portal. Ligament footprint is debrided from distal (a) to proximal (b). The deltoid ligament is observed detached from the bone (c). (1) Medial malleolus. (2) Talar dome. (3) Anterior area of the deltoid ligament

Full size image
Fig. 3
figure 3

Arthroscopic view of the medial gutter in a left ankle. Scope introduced through the anterolateral portal and directed to the medial gutter. Automatic suture passer clamp introduced through the anteromedial portal. The area of the ligament disinserted is penetrated with an automatic suture passer clamp (MiniScorpio, Arthrex, Naples, FL, USA) (a, b). Once the ligament is penetrated by the double suture (c), the ends of the suture are introduced into the suture loop, and by pulling the ends, the loop is introduced into the joint and the ligaments grasped by the suture (d)

Full size image

The appropiate location for the suture anchor was then identified. Optimal placement ought to be close to the insertion of the deltoid ligament, on the anteromedial aspect of the medial malleolus. An anatomical reference for the placement of the anchor is the intersection formed by the ankle joint line and a perpendicular line extending from the lateral border of the tip of the medial malleolus. The anchor was to be aimed inferomedially to this intersection (Fig. 4). The drill guide was placed through the anteromedial portal and centered over the anchor. The drill had to be directed parallel to the plantar plane, and from anterior to posterior aiming medially at approximately 10°–15° to avoid violation of the joint space (Fig. 5). The hole was drilled, and the bone anchor (with the suture passed through the portal) was introduced by impaction (Fig. 6). Tensioning of the suture was adjusted before introduction of the anchor. The deltoid ligament was reattached with the ankle positioned in dorsiflexion and neutral rotation (Fig. 7).

Fig. 4
figure 4

Anterior view of the ankle dissection to show the area where the anchor is inserted. The anatomical reference is the intersection formed by the ankle joint line and a perpendicular line extending from the lateral border of the medial malleolus tip. The anchor is aimed to be placed inferomedially to this intersection

Full size image
Fig. 5
figure 5

Arthroscopic view of a right ankle showing the drill guide placed through the anteromedial portal and centered on the inferomedial aspect of the intersection formed by the ankle joint line and a perpendicular line extending from the lateral border of the medial malleolus tip

Full size image
Fig. 6
figure 6

Through the anteromedial portal, the anchor with the suture is introduced by impaction

Full size image
Fig. 7
figure 7

Arthroscopic view of a deltoid ligament “open book” injury before and after an all-arthroscopic repair. a, b Deltoid ligament tear. Ankle in neutral position (a) and probe showing the area disinserted of the ligament (b). Vision after deltoid ligament reinsertion (c). (1) Medial malleolus. (2) Deltoid ligament tear. (3) Area of the deltoid ligament disinserted. (4) Knotless anchor. (5) Deltoid ligament reinserted

Full size image

Venous thromboprophylaxis was given for 10–15 days following surgery. A removable walking boot was kept on at all times for the first 3–4 weeks, and partial weight bearing was permitted. In cases of chondral or osteochondral injuries, no weight-bearing was allowed for 4–6 weeks. Physical therapy was initiated after the walking boot was removed. Physical therapy entailed active and passive range of motion, gait training, strengthening in ankle dorsiflexion, plantarflexion, eversion and inversion, as well as proprioceptive training with weight bearing.

Statistical analysis

Statistical analysis was performed using SPSS 19 (SPSS Inc., Chicago, IL, USA). Descriptive results were presented as median and range.

Results

Thirteen patients (13 ankles) underwent an arthroscopic all-inside lateral and medial ligament repair with a knotless suture anchor technique, and were all retrospectively recruited during follow-up assessment. They were then followed up for a median time of 35 months (range 18–42 months), with no loss of patients during the study period.

During the arthroscopic procedure, 12 patients were diagnosed with an isolated fibular detachment of the ATFL. In one patient, both the ATFL and CFL were involved. A tear affecting the anterior part of the deltoid ligament was observed in all cases. Although the ligament maintained some of its proximal insertional fibers, a separation was observed between the ligament and the medial malleolus, and a probe could be introduced between the bone and the ligament. The deltoid ligament tear was seen to separate from the medial malleolus when an internal rotational force to the talus was applied. The gap was closed when either the tibio-talar joint was in neutral position, or if the talus externally rotated. Because of this deltoid ligament injury pattern, it has been named an “open book” tear.

Five patients underwent additional ankle arthroscopic procedures. These included: an osteochondral talar lesion debridement and microfracture in two patients, osteophyte resection in one, and posterior endoscopy in two. During posterior endoscopy, an os trigonum was excised in one patient, and some intra-articular posteromedial loose bodies were removed in another patient following exposure of the posterior joint capsule.

At final follow-up, all patients reported subjective improvement in their ankle instability. On clinical examination, the anterior drawer test was negative and no talar tilt was elicited in all patients. To date, none of the patients have required revision ankle surgery. All patients returned to their daily activities without difficulties, except for one patient who continued to experience medial ankle pain. Patients who undertook sporting activities sports returned to their preoperative level of performance without limitation. As a result of their arthroscopic ankle stabilization, five patients (38.5%) experienced ankle plantarflexion deficit compared to the contralateral side, although this was less than 10° in all cases. Two patients (15.4%) had a deficit of 5° or less in ankle dorsiflexion, reporting mild discomfort when squatting or kneeling. Deficit in range of motion was not considered a complication but a consequence of the technique itself.

The median AOFAS score increased from 70 (range 44–77) preoperatively to 100 (range 77–100) at final follow-up.

Discussion

The most important contribution of our study is the description of the first arthroscopic all-inside anatomic repair of the lateral collateral ligament and deltoid ligaments, and the presentation of a positive outcome in a series of patients.

The technique described is indicated for the treatment of rotational instability of the ankle. This anatomic ligamentous repair provides ankle stability, whilst offering all the advantages of an arthroscopic technique.

The arthroscopic findings of this series of patients should help expand the knowledge of RAI as a clinical entity. The description of the described “open book” deltoid injury is consistently observed in our series and is likely to represent the hallmark of RAI, a pathology difficult to diagnose clinically. It is highlighted the importance of having a high index of suspicion in recognizing this injury when performing an arthroscopic ligamentous repair.

RAI is defined as the abnormal increase of talar rotation within the tibiofibular mortise. Patients with CAI can present with deltoid ligament abnormalities [2, 10, 19, 21]. Under weight-bearing conditions, ATFL-deficient ankles demonstrate increased anterior translation, internal rotation, and superior translation of the talus [6]. These biomechanical changes may lead to a deltoid ligament tear. It is estimated that 40% of all patients affected of CAI have a partial deltoid injury [19, 30]. A lower incidence was found in the present study, however, with 13 out of 125 (10.5%) patients with CAI. Due to both the difficulty in diagnosing RAI and the lack of symptoms experienced by some patients in the medial ankle, these patients can be underdiagnosed and the true incidence of the condition potentially underestimated.

The deltoid or medial collateral ligament is an important medial stabilizer of the ankle joint [21, 27, 29]. It is a strong multifascicular ligament. Anatomic descriptions are variable; however, most investigators agree that the deltoid ligament is composed of two layers [3, 15, 16, 22, 25, 28]. The deep layer prevents eversion of the ankle joint, and limits lateral displacement and external rotation of the talus [12, 14, 19, 22, 29].

Where there is chronic deficiency of the lateral collateral ligament, the talus is subjected to ongoing excessive internal rotation and anterior translation. This abnormal motion could lead to an injury to the most anterior component of the deltoid ligament and contribute to RAI. As reported in the current study, the specific injury observed in all cases was an “open book” tear of the deltoid ligament. In this injury, the most anterior component of the deltoid ligament is detached from the medial malleolus whilst its proximal attachments remain intact. Because the deltoid ligament provides stability on external rotation [18, 22, 29], ankles with lateral collateral ligament deficiency and a secondary deltoid ligament injury lack constraint to internal and external rotations. Given the significant role that the deltoid ligament plays in maintaining rotational ankle stability, it should be taken into account in the management of CAI. In addition to the lateral collateral ligament repair or reconstruction, the deltoid ligament should be repaired too when injured. The open technique for combined lateral and medial ligament repair with anchors has been published with good clinical outcomes and high patient satisfaction [5]. It is difficult to ascertain if patients would have had similar outcomes if they were treated only for the lateral ligamentous injury. It was from the authors’ experience that some patients treated with isolated lateral ligamentous repair developed postoperative medial ankle symptoms and residual pain. It was only in recognizing the rotational instability injury pattern could we identify those patients who might have benefited from an additional deltoid repair. A cautious look for deltoid ligamentous injury is advised in every patient undergoing arthroscopic ankle ligamentous repair and the authors believe that any identified injury must be anatomically treated to allow for the best results.

RAI is difficult to assess clinically. Affected patients may present with both lateral and medial ankle symptoms; however, in some cases, patients lack of medial symptoms. In the present study, three patients (23%) reported no symptoms around the medial aspect of the ankle. Medial symptoms can range from significant pain to just vague discomfort. The presence of lateral symptoms (pain, discomfort, or a subjective feeling of instability) is a common clinical presentation in patients with RAI. Because there is no specific clinical test to diagnose RAI, imaging studies are mandatory for the detection of deltoid ligament tears in patients with CAI. A careful interpretation of preoperative MRI is essential when RAI is suspected [20, 31]. Both the superficial and deep components of the deltoid ligament are easy to delineate on MRI [7]. The sensitivity and specificity in detecting deltoid ligament tears with MRI is high [8, 11]. Of note, MRI abnormalities of the deltoid ligament have been observed in 55% [24] to 72% [8] of asymptomatic subjects. The patients included in the study had all experienced instability symptoms that warranted surgical intervention. At the time of arthroscopy, the presence of a deltoid “open book” injury associated with chronic lateral ankle instability was highly suggestive of RAI. Although the MRI demonstrated a chronic deltoid ligament tear, the features of an “open book” injury were not observed in any MRI scan, possibly due to the static nature of this imaging modality. As the injured ligament has to be separated from the bone to demonstrate a ligament tear, the authors expect that an MRI arthrogram might be helpful in detecting this type of injury. This highlights the importance of careful arthroscopic assessment even if the deltoid ligament appears intact on MRI without intra-articular contrast. Arthroscopic evaluation of the deltoid ligament is advisable, before the torn lateral collateral ligament is treated. The ability to introduce arthroscopic instruments between the medial maleollus and the deep layer of the deltoid ligament, and the visualization of the ligament separating from the bone when passively rotating the ankle are the cardinal signs that indicate an “open book” injury of the deltoid ligament. Only a dynamic modality like arthroscopy can provide an accurate diagnosis of this injury.

In the present series, medial ankle pain or discomfort disappeared postoperatively in all but one patient, and was not elicited in those without preoperative medial symptoms. Persistent medial ankle symptoms can be expected when the deltoid injury is left untreated in patients with RAI. It is, therefore, recommended that any deltoid ligament tear be repaired during surgery even in patients without medial symptoms that undergo arthroscopy for CAI. Of note, diminished range of motion—both plantar and dorsiflexion—was observed in some cases following this ligamentous repair. The deficit in ankle plantar flexion (38.5% of patients) was less than 10° and carried no clinical consequences. Literature suggests that this may be a sequel of the ATFL repair itself [35]. Conversely, the deficit seen in dorsiflexion (15.4% of patients) which was less than 5° was perceived by patients as a restriction during active ankle dorsiflexion. Prolonged immobilization could hypothetically cause a dorsiflexion deficit. We would, therefore, recommend restriction of plantarflexion only, and allowing free dorsiflexion during the immobilization period.

Study limitations include the lack of a comparative control group with our study subjects who underwent arthroscopic all-inside lateral collateral ligament repair. The benefit of treating both collateral ligaments is supported by a previous study treating RAI via open surgery with similar satisfactory results [5]. Furthermore, a number of associated injuries were observed and treated in this cohort of patients, potentially introducing a confounder to the effect of the stabilizing procedure. This limitation is a consequence of the natural history of CAI and its typically associated injuries. In essence, it is hardly possible to find a homogenous cohort of isolated CAI without concomitant ankle injuries. Another limitation is that the AOFAS score used is not a validated outcome scale nor a specific item to evaluate ankle instability, and consequently, some clinical aspects may have been overlooked. A specific clinical score to assess ankle instability would have increased the validity of the study.

Conclusion

In conclusion, RAI can be successfully treated arthroscopically. The presence of a deltoid “open book” tear affecting the anterior part of the ligament in patients with CAI should be considered as a sign of RAI.