The Effect of Modified Brostrom-Gould Repair for Lateral Ankle Instability on In Vivo Tibiotalar Kinematics

Abstract

Background

Lateral ankle instability leads to an increased risk of tibiotalar joint osteoarthritis. Previous studies have found abnormal tibiotalar joint motions with lateral ankle instability that may contribute to this increased incidence of osteoarthritis, including increased anterior translation and internal rotation of the talus under weight-bearing loading. Surgical repairs for lateral ankle instability have shown good clinical results, but the effects of repair on in vivo ankle motion are not well understood.

Hypothesis

The modified Broström-Gould lateral ligament reconstruction decreases anterior translation and internal rotation of the talus under in vivo weight-bearing loading conditions.

Study Design

Controlled laboratory study.

Methods

Seven patients underwent modified Brostöm-Gould repair for unilateral lateral ankle instability. Ankle joint kinematics as a function of increasing body weight were studied with magnetic resonance imaging and biplanar fluoroscopy. Tibiotalar kinematics were measured in unstable ankles preoperatively and postoperatively at a mean follow-up of 12 months, as well as in the uninjured contralateral ankles of the same individuals.

Results

Surgical repair resulted in statistically significant decreases in anterior translation of the talus (0.9±0.3mm, p=0.018) at 100% bodyweight and internal rotation of the talus at 75% (2.6±0.8°, p=0.019) and 100% (2.7±0.8°, p=0.013) bodyweight compared to ankle kinematics measured before repair. No statistically significant differences were detected between repaired ankles and contralateral normal ankles.

Conclusion

The modified Broström-Gould repair improved the abnormal joint motion observed in patients with lateral ankle instability, decreasing anterior translation and internal rotation of the talus.

Clinical Relevance

Altered kinematics may contribute to the tibiotalar joint degeneration that occurs with chronic lateral ankle instability. The findings of the current study support the efficacy of this repair in improving the abnormal ankle motion observed in these patients.

INTRODUCTION

Ankle sprains are very common, comprising 14 to 23% of all sports injuries^{17, 18}. The majority of sprains are caused by an inversion mechanism, which accounts for 80% of sprains¹⁷. As a result, the lateral ankle ligaments are most frequently injured, with injury to the anterior talofibular ligament (ATFL) the most common¹⁶. In more severe sprains, ATFL injury may be accompanied by injury to the calcaneofibular ligament (CFL)^{13, 26}. These ligaments are injured in 79 to 85% of ankle sprains^{18, 41}. Conservative treatment for acute ankle sprains, which includes rest, icing, and early protected weight-bearing with bracing, results in good clinical outcomes in most patients^{3, 29, 32}. However, up to 40%^{18, 21, 35, 43} of patients suffering lateral ankle sprain will continue to have symptoms despite appropriate conservative therapy, including pain, subjective instability, recurrent sprains, and proprioceptive deficits³.

In addition to these symptoms, long-term lateral ankle instability may lead to an increased risk of early-onset osteoarthritis. Studies have found that 16 to 79% of patients with untreated lateral ankle instability develop arthritis^{9, 20, 25, 42}. Unlike most major joints, posttraumatic arthritis of the ankle is much more common than primary osteoarthritis, comprising 78% of ankle arthritis⁴². Previous researchers have hypothesized that altered joint mechanics contribute to the development of osteoarthritis^{2, 20–22, 41}, motivating a number of studies investigating changes in joint motion and cartilage contact associated with lateral ankle instability^{31, 33, 36}.

In an effort to restore normal ankle function, there have been many surgical procedures developed for treating lateral ankle instability^{10, 12, 14, 15}. Of these many techniques, those reconstructing the native ligamentous anatomy have shown the best clinical results and fewest complications. Using various outcome scoring systems, good or excellent short-term and long- term clinical results have been reported in 85 to 100% of patients undergoing anatomic lateral ligament reconstruction^{5, 27, 30}. One of these anatomical repair procedures was described by Broström in 1966 and involved direct repair of the lateral ankle ligaments¹⁰. Gould et al later described a modification of this anatomic reconstruction involving mobilization and reattachment of the extensor retinaculum to augment the repair¹⁹.

Despite the excellent clinical outcomes of these anatomic repair procedures, there has been evidence that patients may still have an increased risk of ankle arthritis²³. Several cadaveric studies have analyzed the effects of anatomic repairs on ankle joint mechanics^{4, 24}, but there is very little data on how these repairs affect in vivo ankle joint biomechanics. Thus, the purpose of the current study is to examine how the Broström-Gould repair affects the in vivo kinematics of the tibiotalar joint during quasi-static weight-bearing loading^{8, 11}. We hypothesize that this repair helps to restore the motion of the intact ankle, decreasing the increased anterior translation and internal rotation of the talus observed in a previous study of patients with lateral ankle instability¹¹.

MATERIALS AND METHODS

Patient Recruitment

Patients were recruited from the Foot and Ankle service at our institution with IRB approval. All patients were diagnosed with unilateral, mechanical lateral ankle instability by an orthopaedic surgeon specializing in foot and ankle surgery and planned to undergo modified Broström-Gould repair of the affected ankle^{8, 11}. Three of the patients who enrolled in the present study participated in a previous study examining ankle kinematics prior to repair in our laboratory¹¹. Each patient had a history of an ankle sprain treated non-operatively for at least six months with persistent symptoms of ankle instability and positive physical examination findings of mechanical lateral ankle instability. Symptoms included pain, subjective instability, and/or recurrent sprains. Physical examination included the anterior drawer test and the talar tilt test. Each patient underwent an MRI examination of both ankles to confirm lateral ligament pathology, which was interpreted by a fellowship-trained and board-certified musculoskeletal radiologist. Patients with osteochondral lesions or peroneal tendon injuries detected on MRI or during surgery were excluded. Patients were required to have contralateral normal ankles as confirmed by clinical examination and MRI, so patients with contralateral pathology were also excluded. Patients returned to have their ankles studied 12 months postoperatively.

Surgical Repair

Each patient underwent modified Broström-Gould anatomical reconstruction performed by a board-certified orthopaedic surgeon specializing in foot and ankle surgery. The anterior talofibular ligament was reconstructed in each patient, and the calcaneofibular ligament was reconstructed if CFL pathology was identified on MRI or during surgery.

The procedure used a small incision from the anterior border of the fibula superiorly to the peroneal tendon sheath, curving obliquely toward the sinus tarsi. The extensor retinaculum was elevated and a small anterior arthrotomy was made to inspect the joint surface for osteochondral lesions. The peroneal tendon sheath inferior to the tip of the fibula was exposed, and a small incision was made in the peroneal retinaculum to inspect the peroneal tendons. Once the absence of osteochondral lesions and peroneal tendon pathology was confirmed, the ankle capsule containing the anterior talofibular ligament and, if indicated, the calcaneofibular ligament were incised off the fibula and elevated as a flap. A small distance of periosteum was then elevated off of the anterior aspect of the fibula, and two or three 3.5mm corkscrew suture anchors were then inserted into the fibula as the foot was everted and dorsiflexed. The ATFL and, if indicated, CFL were sutured to the anchors. The repair was further augmented by incorporating the extensor retinaculum. The surgeon confirmed that the patient had a negative anterior drawer and talar tilt before closing the wound.

Imaging

Preoperative MR imaging was performed bilaterally^{8, 11} in the Department of Radiology at our institution. Each ankle was imaged separately in the sagittal plane with a 3.0T magnet (Trio TIM, Siemens, Germany) using a dedicated 8-channel receive-only foot and ankle coil (Invivo, Orlando, Florida). The protocol used a three-dimensional double echo steady state sequence (flip angle: 25°, echo time: 6 ms, repetition time: 17 ms), a 512×512 pixel resolution spanning a 15cm × 15cm field of view, and a slice thickness of 0.7mm. Scanning was performed twice, once with water excitation and once without, to create two series of sagittal images for each ankle.

Solid modeling software (Rhinoceros 4.0, McNeel & Associates, Seattle, Washington) was then used to outline the bony anatomy of the tibia and talus for each ankle. Using both water excitation and non-water excitation images allowed for the delineation of the bone and soft tissue interface in each MRI slice (Figure 1). Each outlined contour was then used to create three-dimensional models of the tibia and talus (Figure 1).

Figure 1.

High resolution MR images were segmented to create a 3D model of the tibia and talus (top row). Next, each subject was imaged using biplanar fluoroscopy while standing with increasing body weight on the foot (bottom left). Finally, the 3D models of the tibia and talus were manipulated in six degrees-of-freedom until their projection matched that of the orthogonal fluoroscopic images to reproduce the in vivo motion of the ankle (bottom right).

Following MRI, each patient underwent biplanar fluoroscopy preoperatively^{8, 11} using two orthogonally positioned fluoroscopes (Pulsera, Philips, Netherlands), each with a resolution of 1024×1024 pixels (Figure 1). Both ankles were imaged separately. The subject’s total bodyweight while wearing lead protection was measured initially. During imaging, the subject stood on a platform with one foot on a forceplate so that the ankle was within the beams of the orthogonal fluoroscopes. One fluoroscope provided an anteromedial view of the ankle and the other an anterolateral view (Figure 1). The contralateral foot was placed behind the subject to provide stability and not interfere with imaging. The subject adjusted his or her weight between the foot on the forceplate and the contralateral leg such that the forceplate display matched the desired load. Images were taken with 25%, 50%, 75%, and 100% bodyweight being placed on the ankle in a neutral standing position.

Twelve months after surgery, each subject returned for another kinematic evaluation. During the postoperative study, the same fluoroscopic imaging procedure was repeated, again creating pairs of orthogonal fluoroscopic images of the patient’s ankles while standing with 25%, 50%, 75%, and 100% bodyweight on each leg.

Kinematic Measurements

Fluoroscopic images were imported into the solid modeling software (Rhinoceros 4.0, McNeel & Associates, Seattle, Washington) so that their position reproduced the orthogonal orientation of the fluoroscopes during the examination (Figure 1). The three-dimensional models of the tibia and talus created from the preoperative MRI images were then imported into this virtual orthogonal fluoroscope environment. Using a custom-written edge detection program (Mathematica 6.0, Wolfram, Champaign, Illinois), gradients in pixel intensity were used to outline the bones in each fluoroscopic image. The tibia and talus models were then individually manipulated in six degrees-of-freedom (three translations and three rotations) so that their projections precisely matched the outlines on the fluoroscopic images as viewed from each of the two orthogonal directions (Figure 1). As a result, these models recreated the in vivo positions of the tibia and talus¹¹.

In order to measure the in vivo kinematics from these models, a Cartesian coordinate system was applied to the three-dimensional models of the tibia and talus¹¹. After using an iterative closest point technique^{1, 6} to align the bones from each patient’s contralateral ankles, coordinate systems were created on both ankles simultaneously¹¹. This process allows for kinematics to be measured on both the intact and injured (before and after repair) ankles using the same coordinate system, thus reducing the variability of kinematic measurements. Based on previously published findings in patients with lateral ankle instability and no repair¹¹, we focused our measurements on the anterior-posterior translation and internal-external rotation of the talus relative to the tibia. Anterior translation was defined as translation of the center of the talar dome along the anterior axis of the tibia, and internal rotation was defined as rotation about a proximal-distal axis fixed to the talus¹¹.

Data Analysis

For each loading condition (25%, 50%, 75%, and 100% bodyweight), the difference in anterior translation and internal rotation of the talus was calculated between the injured ankle before and after repair, as well as the repaired ankle relative to the contralateral intact ankle. The Shaprio-Wilk W test was used to test the data for normality (Statistica, StatSoft, Tulsa, Oklahoma). Paired t-tests and the Bonferroni correction were used to detect statistically significant differences in motion between ankle states. Differences were considered significant where p < 0.025.

RESULTS

Seven patients (three men and four women), aged 37±4 years, fulfilled the inclusion criteria and were studied preoperatively and postoperatively (12±2 months after surgery, mean±sem). On MRI, two patients had isolated anterior talofibular ligament damage, and five patients had damage to both the anterior talofibular ligament and calcaneofibular ligament in the affected ankle.

Comparing anterior translation of the talus in ankles before and after repair, we found a significant decrease in anterior translation of the talus (0.9±0.3mm, p=0.018) after surgery at 100% bodyweight (Table 1). No statistically significant differences were detected between the anterior translation of the talus in repaired ankles compared to contralateral normal ankles (p>0.55, Table 1). At 100% bodyweight, the anterior translation of the repaired ankles was within 0.0±0.3mm of the contralateral normal ankles.

Table 1.

Repair decreased tibiotalar motion relative to that measured before repair. No statistically significant differences in motion were observed between contralateral normal and repaired ankles. Positive values represent an increase in tibiotalar motion, and negative values represent a decrease.

Anterior Translation of Talus (mm)
Body Weight	25%	50%	75%	100%
After repair relative to before repair	0.2±0.9	0.2±0.3	−0.4±0.4	−0.9±0.3*
After repair relative to contralateral	0.5±1.0	0.0±0.5	−0.2±0.3	0.0±0.3
Internal Rotation of Talus (°)
Body Weight	25%	50%	75%	100%
After repair relative to before repair	−1.3±1.2	−0.4±0.9	−2.6±0.8*	−2.7±0.8*
After repair relative to contralateral	0.7±1.8	0.7±2.0	−0.5±1.1	0.7±1.9

^*(p < 0.025, n = 7)

Comparing the internal rotation of the talus in ankles before and after repair (Table 1), we found significant decreases in internal rotation of the talus after surgery at 75% (2.6±0.8°, p=0.019) and 100% bodyweight (2.7±0.8°, p=0.013). Comparing internal rotation of the talus in repaired ankles to contralateral normal ankles, no statistically significant differences were detected (p > 0.71, Table 1). At 100% body weight, the internal rotation of repaired ankles was within 0.7±2.0° of the contralateral normal ankles.

DISCUSSION

Previous studies have established the presence of altered mechanics in ankles with lateral ankle instability as compared to normal ankles^{31, 36}. In an effort to restore normal ankle function after lateral ankle instability, numerous surgical procedures have been proposed^{8, 10, 13, 14}. However, the effects of these procedures on in vivo ankle function are not well understood. Thus, the current study used biplanar fluoroscopy and MR imaging to investigate how the modified Broström-Gould anatomic repair for lateral ankle instability affects the in vivo kinematics of the tibiotalar joint under physiologic, quasi-static loading conditions^{8, 11}.

Our results suggest that the Broström-Gould anatomic repair helps to restore normal motion after lateral ankle instability. A study in a similar population of patients with lateral ankle instability found significant differences in the in vivo kinematics of unstable ankles compared to contralateral normal ankles under quasi-static loading conditions¹¹. In particular, unstable ankles had significantly more anterior translation (1mm) and internal rotation of the talus (6°) at full bodyweight as compared to contralateral normal ankles¹¹. In the present study, similar differences between the intact and injured ankles prior to repair were observed (1mm of anterior translation and 4° of internal rotation). However, Broström-Gould repair significantly decreased the anterior translation of the talus by approximately 1mm at 100% bodyweight and decreased internal rotation of the talus by approximately 3° at 75% and 100% bodyweight.

Many of the reconstruction techniques for lateral ankle instability described in the literature have the goal of providing clinical relief as well as mitigating this increased risk of osteoarthritis^{8, 10, 13, 14}. Thus, restoring native joint motion is an important goal of repair because altered joint mechanics are believed to contribute to the early development of osteoarthritis in patients with lateral ankle instability^{20–22, 41}. Specifically, several previous studies have reported that the abnormal motions that occur with lateral ankle instability alter cartilage loading in the joint^{31, 33, 34}. For example, a previous in vivo study found that the increased anterior translation and internal rotation of the talus resulting from lateral ankle instability increased and caused an anteromedial shift of the peak cartilage contact strain⁸. These abnormal contact strains might alter the mechanical environment of the cartilage and predispose the tibiotalar joint to degenerative changes. Arthroscopic studies have shown that the majority of cartilage lesions in patients with lateral ankle instability are located in the anteromedial portion of the joint^{37, 38}, corresponding to this region of increased strain⁸. Therefore, understanding how surgical repair affects in vivo joint biomechanics after lateral ankle instability is potentially important to improving long-term results.

A number of studies have used cadaveric models to investigate the ability of surgical repair to restore normal ankle biomechanics after sectioning of the lateral ankle ligaments. These studies have shown differing results, which are likely due to the different surgical techniques and loading conditions used. For example, Bahr et al performed a study analyzing the effects of Broström repair on ankle motion⁴. They reported that the Broström repair decreased the increased internal rotation observed with sectioning of the ATFL and CFL, but noted limitations in its ability to restore normal anterior translation under anterior drawer loads⁴. Liu and Baker performed a cadaver study applying an anterior load to the talus²⁴. Using the Broström repair with the Gould modification, they found that repair significantly decreased anterior translation of the talus²⁴. Prisk et al performed a study examining ankle biomechanics under compressive and inversion loading in cadavers with lateral ligament sectioning (both ATFL and CFL) and subsequent Broström repair with and without the Gould modification³⁰. They found that sectioning the ATFL and CFL resulted in a medial shift in the center of pressure of the tibiotalar joint. The Broström repair with the Gould modification decreased this medial shift, while the Broström repair alone did not consistently restore the center of pressure. However, the Broström repair with the Gould modification resulted in increased contact areas compared to the intact ankle. While these studies have contributed important data on the surgical treatment of the lateral ankle ligaments, additional studies are needed to evaluate the ability of various repair techniques to restore normal in vivo joint function and prevent long-term joint degeneration in patients.

In the present study, the decreases in anterior translation and internal rotation of the talus after surgical repair potentially contribute to the excellent clinical outcomes of the modified Broström-Gould procedure. However, despite the encouraging clinical results of both anatomic and non-anatomic repairs, there is suspicion that an increased risk of osteoarthritis remains after repair²³. Krips et al found 59% of patients undergoing Broström repair and 77% of patients undergoing Evans tenodesis, a non-anatomic repair, developed osteoarthritis at a mean follow-up of 19.9 years and 21.8 years after surgery respectively²³. Future studies might evaluate the effects of other repair types on in vivo ankle motion, as well as following these patients long-term to evaluate the efficacy of different repairs in preventing degenerative changes to the joint.

The present study found no statistically significant differences between the kinematics of repaired ankles and normal ankles. The residual differences between repaired ankles and contralateral normal ankles were 0.0±0.3mm anterior translation and 0.7±1.9° internal rotation at 100% bodyweight. These differences are relatively small compared to the differences (approximately 1mm of anterior translation and 6° of internal rotation at 100% bodyweight) between laterally unstable ankles and contralateral ankles before surgery¹¹. A retrospective power analysis can be performed to assess this study’s ability to detect residual differences in anterior translation and internal rotation. Using the observed standard deviations of differences in anterior translation and internal rotation between repaired ankles and contralateral normal ankles, sample sizes of 17 patients or 24 patients would be needed for a power of 0.80 to conclude that there were no differences in kinematics to within 0.5mm or 3°, respectively. These differences represent a 50% correction of the difference in ankle kinematics under full bodyweight found before repair between laterally unstable and contralateral normal ankles by Caputo et al¹¹. While more patients are needed to conclude that this repair restores normal ankle kinematics to within these differences, the statistically significant decrease in motion after repair is encouraging for the efficacy of this technique. Further study is needed to determine what differences in motion are meaningful to long-term outcomes.

This study has some limitations to note. The standing position used is an example of quasi-static physiologic loading. Future studies might quantify dynamic joint motion during more demanding activities^{7, 28, 39, 40, 44} in patients with lateral ankle instability and surgical repair. Furthermore, with only seven patients, the independent contributions of the ATFL and CFL ligaments could not be assessed individually. A future study with a larger sample size would allow us to separately analyze kinematic outcomes in patients with only ATFL injuries and with combined ATFL and CFL injuries. Future studies might also investigate the ability of this repair or other surgical procedures to restore normal in vivo cartilage contact strains⁸.

In conclusion, this in vivo study found that the modified Broström-Gould repair for lateral ankle instability significantly improves the kinematics of the talocrural joint under quasi-static weight-bearing loading conditions. Specifically, repair decreased the anterior translation and internal rotation of the talus, which were significantly increased in patients with lateral ankle instability¹¹. In this group of patients, we found no significant differences in kinematics between repaired ankles and contralateral normal ankles. These findings are consistent with the excellent clinical outcomes of this procedure. Future studies might evaluate the ability of surgical repairs to restore normal joint biomechanics and to prevent degenerative changes with long-term follow-up.

What is Known about the Subject

Previous studies have investigated the effects of anatomical repair on lateral ankle instability in cadaveric models. However, there is little data on the effects of repair on the in vivo kinematics of the ankle joint.

What this Study Adds to Existing Knowledge

This study provides in vivo evidence that the Broström-Gould repair for lateral ankle instability significantly improves ankle joint kinematics back towards normal.