Foot & Ankle International

http://fai.sagepub.com/ Comparison of Gait After Total Ankle Arthroplasty and Ankle Arthrodesis
Robert Flavin, Scott C. Coleman, Shay Tenenbaum and James W. Brodsky Foot Ankle Int 2013 34: 1340 originally published online 13 May 2013 DOI: 10.1177/1071100713490675 The online version of this article can be found at: http://fai.sagepub.com/content/34/10/1340

Published by:
http://www.sagepublications.com

On behalf of:

American Orthopaedic Foot & Ankle Society

Additional services and information for Foot & Ankle International can be found at: Email Alerts: http://fai.sagepub.com/cgi/alerts Subscriptions: http://fai.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Oct 1, 2013 OnlineFirst Version of Record - May 13, 2013 What is This?

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

research-article2013

490675

FAIXXX10.1177/1071100713490675Foot & Ankle InternationalFlavin et al

Comparison of Gait After Total Ankle Arthroplasty and Ankle Arthrodesis

Robert Flavin, MD1, Scott C. Coleman, MS, MBA2, Shay Tenenbaum, MD2,3, and James W. Brodsky, MD2

Foot & Ankle International 34(10) 13401348 The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1071100713490675 fai.sagepub.com

Abstract Background: Prior studies reported improved gait after total ankle arthroplasty and better parameters of gait than those reported in earlier studies of patients after ankle arthrodesis. However, there are very limited data prospectively evaluating the effects on gait after ankle arthroplasty compared with ankle arthrodesis. Controversy remains regarding the relative advantages and disadvantages of these 2 treatments and especially the differences in function between them. Methods: We performed a prospective study involving 28 patients with posttraumatic and primary ankle osteoarthritis and a control group of 14 normal volunteers. We compared gait in 14 patients who had undergone ankle arthrodesis with the gait of 14 patients who had ankle arthroplasty preoperatively and at 1 year postoperatively. Three-dimensional gait analysis was performed with a 12-camera digital-motion capture system. Temporospatial measurements included stride length and cadence. The kinematic parameters that were measured included the sagittal plane range of motion of the ankle and the coronal plane range of motion of the ankle. Double force plates were used to collect kinetic parameters such as ankle coronal and plantar flexiondorsiflexion moments and sagittal plane ankle power. Center of pressure (CoP) and its progression in gait cycle were calculated. Results: Baseline parameters showed comparability among the treatment and control groups. Temporospatial analysis, using time as the main effect, showed that compared with ankle arthrodesis, patients with total ankle arthroplasty had higher walking velocity attributable to both increases in stride length and cadence as well as more normalized first and second rockers of the gait cycle. Kinematic analysis, using time and intervention as the main effects, showed that patients who had ankle arthroplasty had better sagittal dorsiflexion (P = .001), whereas those undergoing ankle arthrodesis had better coronal plane eversion (P = .01). Neither ankle arthrodesis nor arthroplasty altered the CoP progression during stance phase. Total ankle arthroplasty produced a more symmetrical vertical ground reaction force curve, which was closer to that of the controls than was the curve of the ankle arthrodesis group. Conclusions: Patients in both the arthrodesis and arthroplasty groups had significant improvements in various parameters of gait when compared with their own preoperative function. Neither group functioned as well as the normal control subjects. Neither group was superior in every parameter of gait at 1 year postoperatively. However, the data suggest that the major parameters of gait after ankle arthrodesis in deformed ankle arthritis are comparable to gait function after total ankle arthroplasty in nondeformed ankle arthritis. Level of Evidence: Level II, prospective comparative study. Keywords: arthritis, biomechanics, gait studies, outcome studies, Scandinavian total ankle replacement (STAR), total ankle arthroplasty, ankle arthrodesis, fusion

The debate in the literature on the benefits of total ankle arthroplasty (TAA) has evolved since the implantation of the first TAA in 1970.23 The evolution of the TAA, which currently uses both 3-component mobile-bearing designs and 2-component fixed-bearing designs, continues to improve patients gait and function.4,9,17,21,22,27,30,33,43 However, because the rates of survivorship are inferior to those of total hip and total knee arthroplasties, the implantation of the TAA in the younger population still remains a contraindication in the majority of surgeons practices.

1 2

St Vincents University Hospital & UCD, Dublin, Ireland Baylor University Medical Center, Dallas, TX, USA 3 Department of Orthopedic Surgery, Chaim Sheba Medical Center, affiliated with the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel Corresponding Author: James W. Brodsky, MD, Baylor University Medical Center, Human Motion Performance Laboratory 411 N. Washington Avenue, Ste 2100, Dallas, TX 75246. Email: james.brodsky@baylorhealth.edu

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

Flavin et al Ankle arthrodesis or fusion (AF) has therefore remained a more commonly accepted procedure for ankle arthritis in this population.25,26,32,40 The disadvantage of ankle arthrodesis is progression of arthritis in the foot joints distal to the ankle, leading to further joint arthrodesis and foot stiffening,5,7,18,26,32 as well as documented persistent gait abnormalities.2,11,26,40,47 Posttraumatic osteoarthritis (OA) is the cause of more than 70% of ankle arthritis.34,40 Posttraumatic OA results in greater stiffness and subsequently reduced range of motion (ROM),46 and in most cases of OA, the resultant center of rotation of the foot and ankle complex is translated distally to the ankle joint and produces hypermobility in abnormal planes of the hindfoot. The increasing rate of ankle arthritis,1,44 which principally is due to an increasing rate of posttraumatic OA, has led to a growing interest in TAA as an alternative option to arthrodesis due to the increased risk of hindfoot joint degeneration.38,45 However, Wood et al46 demonstrated increased rates of stiffness and pain in patients with posttraumatic OA who underwent TAA. Studies have demonstrated better foot and ankle complex clinical scores, temporospatial parameters, and kinematic and kinetic function following TAA.9,17,30,43 Queen et al31 studied 51 patients who had undergone a fixed-bearing TAA. The authors reported improvements in pain and gait up to 2 years postoperatively and maintenance of ankle ROM. However, there are limited studies analyzing the effectiveness of TAA over arthrodesis for the treatment of ankle arthritis. Piriou et al30 compared the Salto total ankle arthroplasty with ankle arthrodesis and demonstrated a greater improvement in velocity and step length in arthrodesis group, while the TAA group demonstrated greater symmetrical timing of gait and restored normal ground reaction forces.30 However, these results can be debated because of the sampling method and gait analysis used. Cadaver studies have demonstrated that the 3-component TAA, when designed on the tibiotalar anatomy, can simulate near normal ankle kinematics.17,42 However, these studies used constraint, pseudo-static mechanical testing on cadaver ankles under standardized external loading environments that were not comparable to physiological loads. Other studies demonstrated increased hysteresis under external constraint loads.6,16,21 Fluoroscopic studies analyzed the kinematics of mobile bearing in a non-weightbearing environment,21,27 while Leszko et al22 studied gait ROM and demonstrated sagittal ROM of 9.2 degrees and coronal ROM of 2.8 degrees. Indirect comparisons between gait analysis studies analyzing TAA versus control and arthrodesis versus control for ankle arthritis demonstrate superior results for TAA.9,17,20,26,39,40,43 However, to the best of our knowledge there are very limited prospective data comparing the gait of patients who have undergone TAA and those who have undergone ankle arthrodesis. This study

1341 was undertaken to compare the changes in preoperative and postoperative gait between patients with total ankle arthroplasty and those with ankle arthrodesis (fusion).

Materials and Methods

A study was performed involving 14 healthy volunteers and 28 patients with end-stage ankle arthritis, of whom 14 patients underwent TAA and 14 patients underwent AF. The 2 intervention groups (TAA vs AF) were matched on the basis of age, gender, and diagnosis. The study was approved by the institutional review board of the medical center. All patients gave written informed consent to participate in this study. All patients in the arthroplasty group had the Scandinavian total ankle replacement ankle prosthesis (STAR). All patients had diagnosis of posttraumatic ankle osteoarthritis or isolated primary osteoarthritis. Patients with inflammatory arthritis were excluded. The selection of surgical procedure was dependent on age, deformity, and patient preference. The exclusion criteria for TAA were age less than 40 years, deformity greater than 10 degrees in the coronal plane, or a request by the patient for ankle arthrodesis, based on informed consent. Patients with larger coronal deformities were judged at the time not to be candidates for arthroplasty based on studies showing higher complications with such deformities.15,46 The control group consisted of 14 volunteers with no foot and ankle abnormality. All patients underwent weight-bearing ankle radiographs, gait analysis, and pedobarography preoperatively and again 1 year after surgery. Radiographs were measured for coronal and sagittal tibiotalar angles preoperatively and at 1 year postoperatively. Subjects underwent a 3-dimensional gait analysis preoperatively and 1 year after surgery. The gait analysis was performed using a 12-camera motion capture system (Vicon Motion Systems, Oxford, UK). Data were recorded at 100 Hz. A segmented foot model consisting of the tibia/fibula, the hindfoot, the midfoot/forefoot, and the hallux was used.3,19,41 Eleven 6-mm reflective spherical markers were used for each limb. Marker placements on each foot included the lateral, medial, and posterior calcaneus; the base and the head of the first and fifth metatarsals; and the hallux. Additional markers were placed on the tibial tuberosity and the lateral and medial malleoli to define the tibia/ fibula. An Euler sequence of rotations ordered by sagittal, coronal, and then transverse positions were used to calculate hindfoot and ankle angles. After marker placement, a static data collection was made to define each subjects neutral position, which was then used in the walking trials.3,41 Subjects walked across a 10-m capture volume at a selfselected pace several times. A minimum of 20 strides were used for averaging and statistical analysis.

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

1342
Table 1. Patient Demographics. Parameters Age, mean (SD), y Gender, No. (%), male Diagnosis, No. (%) Posttraumatic OAfracture Posttraumatic OAinstability Posttraumatic OAcombined Primary OA
Abbreviation: OA, osteoarthritis; SD, standard deviation.

Foot & Ankle International 34(10)

Ankle Arthrodesis (n = 14) 60.7 (16.3) 5 (63) 6 (43) 3 (21) 2 (14) 3 (21)

Total Ankle Arthroplasty (n = 14) 56.9 (8.6) 3 (38) 11 (79) 0 (0) 2 (14) 1 (7)

P Value .448 .403 .140

Gait data were measured and analyzed in identical fashion for each examination. Temporospatial parameters included velocity, cadence, and step length. Kinematic parameters included coronal and sagittal ROM. Kinetic parameters included maximum coronal and sagittal moments and sagittal peak power. Center of pressure (CoP) progression relative to the entire stance phase and relative to heel strike (HS), heel rise (HR), and toe-off (TO) was calculated and standardized relative to the radiographic measurement on a lateral ankle radiograph between the heel marker and the center of the first metatarsophalangeal joint. The vertical ground reaction force (vGRF) curves were calculated preoperatively and postoperatively and compared with the normal group with regard to symmetry, the first vGRF peak, the minimum vGRF in the plateau, and the second vGRF peak. Pedobarography analysis involved calculating the CoP, which was the average best-fit line between the heel marker and the second toe marker.

significance of both time and intervention as the main effects to be calculated. Interaction is determined by the slopes of the lines formed from the pre- and postoperative data in both the time and intervention data. The null hypothesis is set such that if both lines are parallel, then there is no statistical significance. This method allows a better comparative analysis using the data when time is used as the main effect and when intervention is used as the main effect. It also accounts for any crossover effect as data might improve or worsen over time. This statistical analysis is superior to the Student t test, which does not take account of the crossover effect, which would be normally used.8

Results Baseline Comparison of Groups

The baseline demographic data demonstrated no statistically significant difference between the 2 treatment groups (Table 1). When baseline gait parameters were compared between the intervention groups and the control group, controls performed significantly better in all parameters of gait (Table 2). Time as main effect. With time as the main effect, the TAA group had significantly greater increases in velocity (P = .001) and both of its components, cadence (P = .016) and step length (P = .000), than did the arthrodesis group (AF). The AF group performed significantly better with regard to coronal maximum inversion moment (P = .007) and center of pressure at TO (P = .012). Intervention as main effect. When analyzed with intervention as the main effect, the maximum ankle dorsiflexion was significantly better in the TAA group, (P = .038), while the coronal ROM (P = .001) and maximum coronal inversion moment (P = .005) were significantly better in the AF group. Time plus intervention as main effect.When time and intervention were combined as the main effect, the results

Statistical Analysis
Statistical analysis was performed using STATA 11. Independent sample t tests and chi-square tests were conducted to compare baseline characteristics (age, gender, diagnosis). Preintervention baseline, gait, and kinematic parameters among the 3 groups (AF, TAA, and healthy controls) were compared using 1-way analysis of variance. To assess the impact of the 2 interventions, preintervention and postintervention scores within and between the 2 groups were compared using a repeated-measures analysis of variance. Time (preintervention vs 1 year post intervention) was the variable to compare within each group. Treatment (AF vs TAA) was the variable to compare between the 2 groups. Results are presented as means of dependent variables together with standard deviations. All results were analyzed in 3 ways: with time, with intervention, and with time plus intervention, each as the main effect. To determine an accurate assessment of the benefits of ankle arthrodesis and TAA, the significance of the interaction was calculated. Interaction allows the statistical

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

Flavin et al
Table 2. Baseline Gait Comparison Among Groups. Parameters Velocity, m/s Cadence, steps/min Step length, cm Sagittal ROM, degrees Sagittal max moment, Nm/kg Coronal ROM, degrees Coronal max moment, Nm/kg Sagittal max power, W/kg Center of pressure heel strike, % Center of pressure toe-off, % Center of pressure heel rise, % Control (n = 14), Mean SD 1.28 0.09 108.51 5.00 69.92 5.51 32.2 3.7 1.38 0.13 13.0 3.0 0.13 0.08 1.81 0.51 0.15 0.05 1.05 0.10 0.80 0.10 Ankle Arthrodesis (n = 14), Mean SD 0.88 0.22 106.00 9.72 50.46 12.11 15.2 5.6 0.96 0.38 9.4 4.5 0.06 0.07 0.56 0.49 23.46 9.61 0.94 0.12 0.60 0.15 Total Ankle Arthroplasty (n = 14), Mean SD 0.74 0.28 97.81 16.66 44.69 13.35 15.6 3.8 0.94 0.31 3.0 1.7 0.03 0.05 0.83 0.50 28.13 10.68 0.94 0.12 0.71 0.12

1343

P Valuea .000 .047 .000 .000 .000 .000 .002 .000 .000 .021 .001

Abbreviations: ROM, range of motion; SD, standard deviation. a P value determined by 1-way between-groups ANOVA.

demonstrated a significantly greater increase in ankle dorsiflexion in the TAA group (P = .001) and significantly greater plantar flexion in the AF group (P = .008). Total sagittal ROM was almost but not statistically significant, as a result of the offsetting opposite effects of the dorsiflexion and plantar flexion results. There was a mean 4.1 degrees greater increase postoperatively in the total sagittal ROM in the TAA group, which increased from a mean of 15.6 degrees preoperatively to 19.2 degrees postoperatively, whereas the mean total sagittal ROM in the AF group was unchanged (P = .06). (Table 3), There was significantly greater coronal ROM in eversion (P = .01) and total coronal ROM (P = .039) in the AF group. Coronal eversion ROM decreased in the AF group and increased in the TAA group. Total coronal ROM was essentially unchanged in the AF group but almost doubled in the TAA group (time plus intervention), even though the absolute value of total coronal ROM (9 vs 6 degrees) was greater in the AF group (intervention). The statistically significant difference in the preoperative coronal radiographic angle (Table 3) demonstrates the greater degree of coronal plane deformity and the greater magnitude of coronal plane radiographic correction in the AF group than in the TAA group (P = .041), since the final mean values were so similar (no significant difference by intervention). This is expressed by the large SD values in the AF group.

Vertical Ground Reaction Force

The vGRF curve (Figure 1) demonstrates similar flattening of the curve in both treatment groups preoperatively. The first vGRF peak nonsignificantly increased by 7.1% and 4.7% in the TAA and AF groups, respectively, post intervention (P = .675). The minimum vGRF in the plateau nonsignificantly decreased by 4.3% and 1.8% in the TAA and AF groups, respectively (P = .713), and the second vGRF peak nonsignificantly increased by 8.6% and 5.7% in the TAA and AF groups, respectively (P = .544). Although the differences in vGRF were not statistically significant when preoperative and postoperative values were compared at 3 discrete points (HS, TO, HR) along the path of the stance phase, the qualitative differences in the graph are meaningful. The TAA group produced a more symmetrical curve, which was closer to that of the controls than was the curve of the AF group.

Discussion
While ankle arthrodesis has traditionally been the gold standard for the treatment of ankle arthritis, increasing data show early and midterm favorable clinical results with TAA, and there is increasing acceptance of TAA as an alternative to AF in many, although not all, patients with ankle arthritis.12-14,24,30,36,37 The aim of TAA is to reduce pain and to maintain or increase ROM to allow for better function and reduce the long-term consequences on adjacent joints. Since the first TAA was implanted in 1970, studies have used ankle arthrodesis for comparison. To the best of our knowledge, only 2 studies have directly compared gait function between arthrodesis and arthroplasty. Piriou et al30 compared the Salto TAA with ankle arthrodesis in 24 patients, with 12 patients in each group. This study showed that

Center of Pressure
When time plus intervention was used as the main effect, there were no statistically significant differences in the CoP progression at HS (P = .719), HR (P = .096), or TO (P = .072), demonstrating that neither ankle arthrodesis nor TAA altered the CoP progression during stance phase.

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

1344
Table 3. Gait Results and Analysis of Variance. Intervention Parameters Velocity, m/s Time Period Ankle Arthrodesis, Total Ankle Arthroplasty, Mean SD Mean SD 0.88 0.22 0.99 0.21 106.00 9.72 108.05 9.15 50.46 12.11 55.12 9.61 8.3 7.9 1.9 3.9 6.9 8.4 13.1 3.8 15.2 5.6 15.0 2.2 0.96 0.38 1.11 0.32 12.2 16.3 6.2 9.9 2.8 15.3 2.8 9.4 9.4 4.5 9.0 3.3 0.06 0.07 0.13 0.06 0.56 0.49 0.66 0.34 23.46 9.61 18.68 2.82 0.94 0.12 0.99 0.17 0.60 0.15 0.68 0.11 88.9 12.9 89.1 5.1 89.4 6.3 84.6 4.9 0.74 0.28 0.98 0.13 97.81 16.66 108.96 5.06 44.69 13.35 56.01 7.00 6.3 5.0 11.2 6.0 9.3 5.4 8.0 4.3 15.6 3.8 19.2 6.0 0.94 0.31 1.08 0.32 1.7 1.3 5.2 7.2 1.3 1.4 0.8 4.3 3.0 1.7 6.0 4.8 0.03 0.05 0.06 0.06 0.83 0.50 0.84 0.75 28.13 10.68 23.59 4.94 0.94 0.12 0.97 0.10 0.71 0.12 0.68 0.07 89.6 4.1 88.4 2.8 87.3 4.3 86.4 4.3 Time .001 .016 .000 .614 .072 .080 .084 .476 .133 .099 .007 .700 .773 .012 .054 .325 .133

Foot & Ankle International 34(10)

P Value Intervention .261 .275 .505 .038 .439 .125 .827 .107 .725 .001 0.005 0.134 1.000 0.929 0.013 0.140 .803 Time Intervention .167 .087 .095 .001 .008 .060 .929 .010 .074 .039 .267 .772 .719 .072 .960 .041 .889

Pre Post Cadence, steps/min Pre Post Step length, cm Pre Post Sagittal ROM Pre dorsiflexion, degrees Post Sagittal ROMplantar Pre flexion, degrees Post Sagittal ROM, degrees Pre Post Sagittal max moment, Pre Nm/kg Post Coronal ROMeversion, Pre degrees Post Coronal ROMinversion, Pre degrees Post Coronal ROM, degrees Pre Post Coronal max moment, Pre Nm/kg Post Sagittal max power, W/ Pre kg Post Center of pressure heel Pre strike, % Post Center of pressure toe- Pre off, % Post Center of pressure heel Pre rise, % Post Coronal radiographic Pre angle, degrees Post Sagittal radiographic Pre angle, degrees Post

neither intervention restored patients to normal gait analysis parameters but that the ankle arthrodesis group had better velocity, greater step length, and greater asymmetry in gait pattern, whereas the TAA group showed greater joint movement, symmetry in gait, and more normal restoration of the pattern of ground reaction force. However, the limitations of this study were the sampling method, which resulted in different TAA patients being analyzed at 6 months and 1 year, and the lack of preoperative gait analysis data for the arthrodesis group, therefore excluding baseline comparisons to ensure comparability of the 2 groups. Hahn et al14 compared Salto TAA with ankle arthrodesis in 18 patients (9 patients in each group) without a normal control group. This study showed improved temporospatial

measurements but without significant difference between the groups. Furthermore, arthrodesis patients exhibited a greater increase in hip ROM whereas ankle arthroplasty patients had a greater increase in both knee and ankle ROM and a trend of increasing ankle power generation during late stance. The authors noted that both treatments resulted in decreased pain and improved temporospatial measures of gait, with high plantar flexor moment in the arthrodesis group and better ROM in the arthroplasty group. The Hahn et al study, which is the most comparable study published to date, had the same selection bias of patients assigned to interventions as our study. Our study was undertaken to better compare the effect on gait of ankle replacement versus ankle arthrodesis and to

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

Flavin et al

1345

Figure 1. Vertical ground reaction force curves comparing the pre and postoperative TAA and AF groups to the controls. AF, ankle fusion; STAR, Scandinavian total ankle replacement.

delve into those comparisons in greater detail, using the prospective gait data, as well as to compare patients undergoing these interventions with matched normal controls.

Statistical Methods and Analysis of Results

Our study demonstrated preoperative comparability of the 2 study groups. There was no preoperative difference in the temporospatial parameters and kinetic gait analysis between the TAA and arthrodesis groups except for a statistically significant increase in the coronal ROM of the foot and ankle complex of the arthrodesis group when compared with the TAA group, which correlated with the differences in the preoperative coronal angle, that is, the greater coronal plane deformity in the arthrodesis group. In the statistical analysis, if only time had been used as the main effect (ie, comparison of preoperative and postoperative values between the replacement and arthrodesis groups), this could have led to inaccurate statistical analysis because the patients were not randomly assigned to an intervention and the groups differed with regard to the selected surgical intervention. Thus, intervention (replacement vs arthrodesis) must be then considered distinctly in order to accurately analyze the data, for which ANOVA was required. The results were analyzed separately with reference to time, with reference to intervention, and with reference to time plus intervention combined (Table 3). This is very different than the statistical methods most often used in orthopedic studies, that is, comparison of preoperative and postoperative mean values within a group or between 2 groups of patients.

The purpose of these 3 sets of ANOVA comparisons (see last 3 columns of Table 3) is to afford a more specific and accurate understanding of the comparisons of results. The statistical analysis using time as the main effect expresses what is done in most statistical analyses in the orthopedic literature, that is, comparison of the magnitude of change from preoperative to postoperative values between 2 groups. This would be expressed as a statistically greater increase in, for example, velocity in TAA patients (0.74 0.98m/s = delta of 0.24) than AF patients (0.88 0.99 m/s = delta of 0.11), which was statistically significant (P < .001). In addition, the statistical analysis using intervention as the main effect compares the postoperative results between the 2 groups to evaluate the difference in final effect. Thus, the final velocity in TAA patients of 0.98 m/s and in AF patients of 0.99 m/s (delta of 0.01 m/s) was not statistically significant (P = .261). Finally, the statistical analysis using time and intervention combined gives the truest and most accurate representation of statistically significant differences of the effect of the intervention by comparing all the results between the groups while also comparing each result before and after surgery. In the case of velocity, there is no significant difference between the TAA and AF groups (P = .167). At times, simple comparison of a single aggregated mean can obscure an accurate interpretation of the data, especially if there is a wide range within the group. This is best revealed by the magnitude of the standard deviations. This is shown, for example, in the values of coronal angulation (Table 3). The difference in changes before and after surgery between the AF (88.9 89.1) and TAA patients

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

1346 (89.6 88.4) is not significant (time as main effect). The difference in final values of 89.1 versus 88.4 is also not statistically significant (intervention as main effect). However, the time plus intervention is statistically significant, because the range of severity of coronal deformity in the AF group is much greater than in the TAA group (preoperative SD of 12.9 in AF vs 4.1 in TAA, and postoperative SD of 5.1 in AF vs 2.9 in TAA), meaning that the effect of averaging when using only time or only intervention obscured the differences between the groups.

Foot & Ankle International 34(10) further corroborated the coronal ROM data in reinforcing the clinical observation of increased motion and function in the AF patients because they had greater preoperative coronal deformity at the ankle. Overall, the data showed that the ankle arthritis patients in this study preserved sagittal plane dorsiflexion and hindfoot eversion over other foot and ankle motions. The nonsignificant statistical difference in CoP indicated that ankle abnormality alone was not a factor in CoP progression.

Sagittal Plane ROM

Surprisingly, not all the ROM data were superior in the TAA group. The TAA group had greater gain in ankle dorsiflexion, but the AF group had great plantar flexion. The total sagittal ROM trended higher in the TAA group (P = .06) but failed to reach statistical significance, perhaps due to sample size. These data indicate that there is already compensatory hypermobility in the sagittal plane, most likely at the talonavicular joint, in the arthritis patients preoperatively and that it continues to be expressed in gait postoperatively. The capability of the TAA to alter gait significantly in the sagittal plane may be impeded by compensatory mechanisms that occur in the foot and ankle in response to ankle stiffness, which leads to hypermobility in adjacent joints such as the talonavicular and subtalar joints.10,18,28,29,35 This hypermobility may have 2 effects. First, it may contribute to a more symmetrical and efficient gait cycle by substituting for loss of motion at the ankle. But it also has been postulated as the reason for accelerated hindfoot joint degeneration, although this has yet to be definitively proven in the literature.10,18,28,29,35 Second, the hypermobility may cause the hindfoot complex to less effectively lock the Chopart joint in the heel rise phase of the gait cycle, therefore reducing the need for ROM of the ankle joint, and, in essence, to substitute for ankle ROM. This is suggested by the sagittal ROM that was maintained in the AF group (Table 3).

Vertical Ground Reaction Force

When comparing the vGRF patterns preoperatively and postoperatively, we demonstrated that both groups showed similar preoperative abnormalities and both groups produced similar improvements postoperatively compared with themselves and compared with the control group. This indicated that the increased mobility in the hindfoot joints played an important role in maintaining the pattern of gait, especially in the AF patients. Our results showed specific parameters of gait that were better in the arthrodesis group and some parameters better in the arthroplasty group. The overall evaluation, at 1 year, did not clearly demonstrate complete superiority of one treatment over the other, although the absence of statistically significant difference is not equivalent to the presence of statistically significant sameness. However, one of the most important findings of this study was that it showed comparability of early functional results between patients with minimal deformity who underwent ankle arthroplasty and patients with great ankle deformity who underwent arthrodesis. We expected that some parameters of gait would consistently be different, such as sagittal plane ROM in the arthroplasty patients, whose tibiotalar joint motion was preserved. However, the differences in that parameter were not as great as expected. In the arthrodesis patients, we interpreted this to be due to the compensatory hypermobility of the talonavicular joint, whose sagittal plane motion was indistinguishable from that of the tibiotalar joint using current gait analysis techniques, even with segmental foot models.

Coronal ROM
Coronal eversion ROM decreased in the AF group, due to correction of eversion (valgus) deformity and translation of the total coronal ROM in the AF group to a less everted (less valgus) position. Total coronal ROM was unchanged in the AF group but doubled in the TAA group. The latter was most likely attributable to additional coronal motion through the arthroplasty itself. Absolute total coronal ROM was greater in the AF group. Surgical intervention resulted in doubling of the coronal moment in both intervention groups, although the absolute value in the AF group was twice that of the TAA group. This

Conclusion
To the best of our knowledge, this is the first detailed and comprehensive prospective study comparing TAA to AF using a 3-component TAA design. It demonstrated differences between each group that have could have advantages and disadvantages on gait in the long-term. However, this small and preliminary study does not yet show evidence of consistently superior gait function in TAA compared with AF. Much work remains to be done in this area, including the need for longer follow-up, greater numbers of patients,

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

Flavin et al and further investigations regarding the differences among patients with ankle arthritis in terms of preoperative ankle stiffness, preoperative deformity, and the presence of compensatory hypermobility in the adjacent hindfoot joints. Acknowledgment
The authors acknowledge and express their gratitude to Dr Jane Whelan for her expert advice and assistance in the statistical analysis, which was paramount to the understanding of this large data set.

1347
13. Haddad SL, Coetzee JC, Estok R, et al. Intermediate and long term outcomes of total ankle arthroplasty and ankle arthrodesis: a systematic review of the literature. J Bone Joint Surg Am. 2007;89:1899-1905. 14. Hahn ME, Wright ES, Segal AD, et al. Comparative gait anal ysis of ankle arthrodesis and arthroplasty: initial findings of a prospective study. Foot Ankle Int. 2012;33:282-289. 15. Haskell A, Mann RA. Ankle arthroplasty with preoperative coronal plane deformity: short-term results. Clin Orthop Relat Res. 2004;(424):98-103. 16. Helm R, Stevens J. Long-term results of total ankle replace ment. J Arthroplasty. 1986;1:271-277. 17. Ingrosso S, Benedetti MG, Leardini A, et al. GAIT analysis in patients operated with a novel total ankle prosthesis. Gait Posture. 2009;30:132-137. 18. Jackson A, Glasgow M. Tarsal hypermobility after ankle fusionfact or fiction? J Bone Joint Surg Br. 1979;61-B: 470-473. 19. Kidder SM, Abuzzahab FS Jr, Harris GF, Johnson JE. A sys tem for the analysis of foot and ankle kinematics during gait. IEEE Trans Rehabil Eng. 1996;4:25-32. 20. King HA, Watkins TB Jr, Samuelson KM. Analysis of foot position in ankle arthrodesis and its influence on gait. Foot Ankle. 1980;1:44-49. 21. Komistek RD, Stiehl JB, Buechel FF, Northcut EJ, Hajner ME. A determination of ankle kinematics using fluoroscopy. Foot Ankle Int. 2000;21:343-350. 22. Leszko F, Komistek RD, Mahfouz MR, et al. In vivo kine matics of the salto total ankle prosthesis. Foot Ankle Int. 2008;29:1117-1125. 23. Lord G, Marotte JH: Total ankle prosthesis: technic and 1st results: apropos of 12 cases. Rev Chir Orthop Reparatrice Appar Mot. 1973;59:139-151. 24. Mann JA, Mann RA, Horton E. STARTM ankle: long-term results. Foot Ankle Int. 2011;32:473-484. 25. Mann RA, Rongstad KM. Arthrodesis of the ankle: a critical analysis. Foot Ankle Int. 1998;19:3-9. 26. Mazur JM, Schwartz E, Simon SR. Ankle arthrodesis: long term follow-up with gait analysis. J Bone Joint Surg Am. 1979;61:964-975. 27. Michelson JD, Schmidt GR, Mizel MS. Kinematics of a total arthroplasty of the ankle: comparison to normal ankle motion. Foot Ankle Int. 2000;21:278-284. 28. Morgan CD, Henke JA, Bailey RW, Kaufer H. Long-term results of tibiotalar arthrodesis. J Bone Joint Surg Am. 1985;67:546-550. 29. Morrey BF, Wiedeman GP Jr. Complications and long-term results of ankle arthrodeses following trauma. J Bone Joint Surg Am. 1980;62:777-784. 30. Piriou P, Culpan P, Mullins M, et al. Ankle replacement ver sus arthrodesis: a comparative gait analysis study. Foot Ankle Int. 2008;29:3-9. 31. Queen RM, De Biassio JC, Butler RJ, et al. J. Leonard Goldner Award 2011: changes in pain, function, and gait mechanics two years following total ankle arthroplasty performed with two modern fixed-bearing prostheses. Foot Ankle Int. 2012;33:535-542. 32. Said E, Hunka L, Siller TN. Where ankle fusion stands today. J Bone Joint Surg Br. 1978;60-B:211-214.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.

References
1. Baumhauer JF, Alosa DM, Renstrom AF, Trevino S, Beynnon B. A prospective study of ankle injury risk factors. Am J Sports Med. 1995;23:564-570. 2. Beyaert C, Sirveaux F, Paysant J, Mole D, Andre JM. The effect of tibio-talar arthrodesis on foot kinematics and ground reaction force progression during walking. Gait Posture. 2004;20:84-91. 3. Brodsky JW, Charlick DA, Coleman SC, Pollo FE, Royer CT. Hindfoot motion following reconstruction for posterior tibial tendon dysfunction. Foot Ankle Int. 2009;30:613-618. 4. Brodsky JW, Polo FE, Coleman SC, Bruck N. Changes in gait following the Scandinavian Total Ankle Replacement. J Bone Joint Surg Am. 2011;93:1890-1896. 5. Buck P, Morrey BF, Chao EY. The optimum position of arthrodesis of the ankle: a gait study of the knee and ankle. J Bone Joint Surg Am. 1987;69:1052-1062. 6. Burge P, Evans M. Effect of surface replacement arthroplasty on stability of the ankle. Foot Ankle. 1986;7:10-17. 7. Coester LM, Saltzman CL, Leupold J, Pontarelli W. Longterm results following ankle arthrodesis for post-traumatic arthritis. J Bone Joint Surg Am. 2001;83-A:219-228. 8. Dupont WD. Statistical Modeling for Biomedical Researchers: A Simple Introduction to the Analysis of Complex Data. Cambridge, UK: Cambridge University Press; 2002. 9. Dyrby C, Chou LB, Andriacchi TP, Mann RA. Functional evaluation of the Scandinavian Total Ankle Replacement. Foot Ankle Int. 2004;25:377-381. 10. Fuchs S, Sandmann C, Skwara A, Chylarecki C. Quality of life 20 years after arthrodesis of the ankle: a study of adjacent joints. J Bone Joint Surg Br. 2003;85:994-998. 11. Fuentes-Sanz A, Moya-Angeler J, Lopez-Oliva F, Forriol F. Clinical outcome and gait analysis of ankle arthrodesis. Foot Ankle Int. 2012;33:819-827. 12. Gougoulias N, Khanna A, Maffulli N. How successful are current ankle replacements? A systematic review of the literature. Clin Orthop Relat Res. 2010;(468):199-208.

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013

1348
33. Saltzman CL, Mann RA, Ahrens JE, et al. Prospective con trolled trial of STAR total ankle replacement versus ankle fusion: initial results. Foot Ankle Int. 2009;30:579-596. 34. Saltzman CL, Zimmerman MB, ORourke M, et al. Impact of comorbidities on the measurement of health in patients with ankle osteoarthritis. J Bone Joint Surg Am. 2006;88: 2366-2372. 35. Schaap EJ, Huy J, Tonino AJ. Long-term results of arthrod esis of the ankle. Int Orthop. 1990;14:9-12. 36. Schuh R, Hofstaetter J, Krismer M, et al. Total ankle arthro plasty versus ankle arthrodesis: comparison of sports, recreational activities and functional outcome. Int Orthop. 2012;36:1207-1214. 37. Slobogean GP, Younger A, Apostle KL, et al. Preference based quality of life of end-stage ankle arthritis treated with arthroplasty or arthrodesis. Foot Ankle Int. 2010;31: 563-566. 38. SooHoo NF, Zingmond DS, Ko CY. Comparison of reopera tion rates following ankle arthrodesis and total ankle arthroplasty. J Bone Joint Surg Am. 2007;89:2143-2149. 39. Stauffer RN, Segal NM. Total ankle arthroplasty: four years experience. Clin Orthop Relat Res. 1981;(160):217-221.

Foot & Ankle International 34(10)

40. Thomas R, Daniels TR, Parker K. Gait analysis and func tional outcomes following ankle arthrodesis for isolated ankle arthritis. J Bone Joint Surg Am. 2006;88:526-535. 41. Tulchin K, Orendurff M, Karol L. A comparison of multi-seg ment foot kinematics during level overground and treadmill walking. Gait Posture. 2010;31:104-108. 42. Valderrabano V, Hintermann B, Nigg BM, Stefanyshyn D, Stergiou P. Kinematic changes after fusion and total replacement of the ankle, part 1: range of motion. Foot Ankle Int. 2003;24:881-887. 43. Valderrabano V, Nigg BM, von Tscharner V, et al. Gait anal ysis in ankle osteoarthritis and total ankle replacement. Clin Biomech (Bristol, Avon). 2007;22:894-904. 44. Valtola A, Honkanen R, Kroger H, et al. Lifestyle and other factors predict ankle fractures in perimenopausal women: a population-based prospective cohort study. Bone. 2002;30:238-242. 45. Winson IG, Robinson DE, Allen PE. Arthroscopic ankle arthrodesis. J Bone Joint Surg Br. 2005;87:343-347. 46. Wood PL, Deakin S. Total ankle replacement: the results in 200 ankles. J Bone Joint Surg Br. 2003;85:334-341. 47. Wu WL, Su FC, Cheng YM, et al. Gait analysis after ankle arthrodesis. Gait Posture. 2000;11:54-61.

Downloaded from fai.sagepub.com at American Orthopaedic Foot & Ankle Society on November 28, 2013