Intraoperative Measurement of Distraction for Ligament Tensioning in Total Ankle Arthroplasty

Correct periarticular ligament tensioning during total ankle replacement seems to be essential to postoperative ankle stability. Lack of appropriate ligament tensioning can be responsible for severely altered joint mechanics and may be a factor in bony impingement.7 Appropriate ligament tension can be achieved either by pretensioning the ligaments before bone resection and implantation, or by adjusting ligament tension after implantation through insertion of the varying thicknesses of bearings or shims. The suggested method for implanting the AgilityTM Total Ankle (Depuy Orthopaedics, Warsaw, IN) includes distraction of the ankle using an external fixator.2 The external fixator serves two purposes. First, it is a means of stabilizing the position of the talus relative to the tibia during bone resection, thereby allowing alignment and deformity correction. Second, distraction of the joint allows the ligaments to be tightened before implant insertion. The distraction force applied during surgery influences the final tension across the implanted components and, as a result, affects the stability and ROM of the ankle. Unfortunately, the proper tension to be applied during distraction of the ankle before component implantation is currently a subjective matter of feel and experience. Current practice recommends distraction to a point of sharply increasing resistance, thought to occur at the maximum deltoid ligament elongation. The optimal distractive tension is as yet unquantified, and no method for measuring it intraoperatively is currently available. A tactile sense of the torque being applied to the distractor, and the visually observed amount of joint opening, are the only feedback available to the surgeon. Without quantifiable guidelines, the skills needed for achieving optimal ligament tension during replacement of the Agility Total Ankle are not transferred easily to surgeons who are new to, or less experienced with, the implantation procedure. Only two previous studies have addressed the question of ligament strain during distraction of the ankle,1,3 both of them in the context of arthroscopic examinations. Based on inspection of eight specimens from cadavers having been subjected to ″excessive distraction″ using a common percutaneous pin distraction device, Guhl3 concluded that the ankle could be distracted safely 7 to 8 mm, for 45 to 60 minutes. No quantitative data were recorded. That study additionally reported clinical results of distraction applied to 84 patients having arthroscopy, finding no complications related to ligament strain. Albert et al1 did a cadaver study to characterize the effects of distraction, recording ligament strain and joint opening as a function of distraction force. They suggested that a maximum tensile load of no more than 135 N be applied across the ankle, which corresponded to an average opening of just under 5 mm. At these loads, they reported no evidence of ligamentous injury attributable to distraction of the foot in the neutral or plantar flexed positions, but they did find evidence of damage to the calcaneofibular ligament when the foot was distracted while in the dorsiflexed position. That result was, however, arguably specific to the type of distractor used, an Acuflex small-joint distractor, with one pin placed each in the calcaneus and distal tibia. Use of a two-pin distractor restricted that study to relatively low distraction forces, because of appreciable pin bending and bony destruction at the site of pin insertion when forces exceeded 135 N. Moreover, testing of normal ankles from cadavers may be of limited value because it does not account for alterations in soft tissue because of disease, chronic disuse, and/or aging normally seen in candidates for ankle replacement. Contemporary total ankle replacement using the Agility implant generally have used an Orthofix (Orthofix SRL, Verona, Italy) or EBI (EBI LP, Parsippany, NJ) distraction device. With these devices, at least three factors contribute to a surgeon’s ability to achieve a given amount of ligamentous tensioning. These factors are: (1) the surgeon’s ability to sense and apply torque of a specified magnitude; (2) the interdevice (distraction) consistency of the relationship between the applied distractor torque and the force of distraction actually achieved; and (3) differences in ankle stiffness attributable to variability in pin placement, quality of pin anchorage, and soft tissue alterations. Therefore, several questions arise regarding the distraction technique. How much distraction force do surgeons currently apply? Is this force reproducible, and what factors might affect its reproducibility? Is resistance to distraction before bone resection an appropriate guide to proper ligamentous tensioning? What relationship exists between the force sensed by the distractor and the force actually applied to ligament tensioning? Does distraction length equate to joint opening? Is ligament tautness maintained after implantation? We sought to develop and implement a device to measure forces applied to distract the ankle and the resulting distraction during total ankle arthroplasty. From these data, we expected to answer or clarify the previously posed questions about ligament tensioning. Force, rather than distraction length, was assumed to be the appropriate guide for proper tensioning of the ligaments. It was hypothesized that the amount of distractive force applied to the joint could be controlled effectively through the surgeon’s tactile sense of torque applied to the distractor unit. Because the long-term success rate has tended to be highest for those prostheses implanted by the most experienced surgeons, we opted to measure the distractive forces applied by the originator and most experienced surgeon with this procedure (FGA). We did so with the intention of eventually using these benchmark data as a teaching guide for surgeons new to the technique. In conjunction with this, we studied the specific distractor used in the originator’s own surgical practice, to understand possible reasons for trial-to-trial variation in the measurements. MATERIALS AND METHODS Use of External Fixator for Distraction The distraction system used in this study incorporates an external fixator (Dynamic Axial Fixator, Orthofix) and a detachable distractor unit (Compression/Distraction Unit, Orthofix) as shown in Figure 1. When assembled, distraction is accomplished by turning the screw of the distractor using a hex head wrench. This lengthens the external fixator at its telescoping segment, thereby distracting the joint.Fig 1.: The photograph shows the Orthofix external fixator, with the distractor unit mounted medially for ankle distraction.The external fixator is applied medially, with one pin each in the calcaneus and neck of the talus,5 and at least two pins in the distal midtibia. Correction of varus or valgus deformity at the ankle is accomplished first using the external fixator. Once corrected, the alignment is locked. The distractor unit then is mounted onto the external fixator, and initial distraction is applied to lightly tension the joint. An anterior approach to the ankle is made, the capsule is opened, and the joint is exposed sufficiently for later bone resection using the cutting guide. Before bone resection, the ankle is distracted again to what the surgeon considers a maximum safe distraction force, or until it is no longer possible to distract further using only finger pressure to turn the wrench. The distractor tension then is released by retreating approximately 2 mm, to allow for the combined 2 mm of porous coating covering the bone ingrowth surface of the implants. All components of the implant then are inserted simultaneously. When the external distraction finally is removed, the tension in the ligaments ideally should be maintained entirely by intercomponent loading of the implant. Instrumentation and Calibration Orthofix distractors were instrumented with strain gauges (Fig 2), in a Wheatstone bridge configuration that allowed the measurement of axial forces in the distractor while eliminating bending effects. A linear potentiometer (Greenpot, Midori Precisions Co Ltd, Tokyo, Japan) was used to measure the distractor travel. Three distractors were instrumented in this manner, to allow for successive ankle replacement procedures on the same day. An MTS (MTS Systems Corporation, Eden Prairie, MN) biaxial materials testing machine was used to calibrate the instrumented distractors, and to evaluate the frictional resistance to distraction inherent to the distractor and the external fixator.Fig 2.: The photograph shows the distractor unit instrumented with strain gauges for force measurement, and with linear potentiometer for distraction length measurement.Calibration of each instrumented distractor was done with the distractor mounted to an external fixator, and with the recorded load applied through the axis of the fixator. Torque then was applied to distract the unit, and the resulting strain gauge output was recorded. This output was approximately linearly proportional to the force recorded by the MTS load cell, yielding a linear conversion coefficient appropriate for each distractor unit. This calibration included the internal static frictional resistance of the fixator. Separately, the relationship between distraction force and applied torque was evaluated for each distractor alone. Cadaveric Trials An unembalmed lower leg was mounted with a fixator and instrumented distractor, following the same procedure as used intraoperatively. The ankle was distracted multiple times, with force and distraction distance being recorded. The gap opening then was inspected for uniformity of distraction. Three additional unembalmed lower legs were used to validate the range of stiffness values measured intraoperatively. Pins were inserted in the calcaneus, talus, and tibia as would be done for intraoperative ankle distraction using an external fixator. No additional incisions or surgical alterations were made before testing. The pins were held rigidly by custom fixtures adapted to the crosshead actuator and load cell of an MTS MiniBionix testing machine. Controlled distraction of the ankle (6 mm/minute) was achieved by applying tension across the joint to a maximum magnitude of 600 N. Apparent joint stiffness was calculated from load versus relative movement apart of the tibia and talus as measured by actuator displacement. Intraoperative Measurement Human subjects approval was obtained, and after sterilization, the instrumented distractors were used intraoperatively on six patients having total ankle replacement for painful degenerative joint disease. Raw data were captured using battery-powered signal conditioning hardware (SCXI system, National Instruments Corporation, Austin, TX) and a laptop computer, allowing online processing and real-time display of force and displacement values during surgery. All ankle replacements were done by one surgeon (FGA), the developer of the implant and originator of the distraction technique. Intraoperative radiographs were taken and later used to evaluate the actual distraction achieved at the ankle before resection for implantation. Distractor force and excursion were recorded during the entire total ankle procedure. The specific details of each procedure varied somewhat from patient to patient, but followed the general order described above regarding external fixator use. In addition to the intraoperative distraction steps typically done, this investigational patient series included several repeat distraction trials taken to the perceived distraction load limit. These additional trials were included to assess repeatability of the joint stiffness measurement and of the surgeon’s tactile feel for the perceived load limit. The study design was deliberately observational, in that the goal was to monitor the loads and joint distractions typically achieved by an experienced surgeon, following his habitual procedure. Even though that procedure was variable, we specifically sought to avoid having the surgeon alter his procedure for purposes of consistency or convenience of data collection. RESULTS In general, the forces resisting distraction, as measured intraoperatively, increased steadily as distraction was applied, revealing a nearly linear relationship between force and distractor excursion in the ankle (Fig 3). Distraction was accommodated uneventfully until a distinct manuallyperceived limit was reached, identified by a perceived difficulty of incrementally distracting further. The force measured just before reaching this perceived limit was fairly reproducible for each ankle, but varied greatly between ankles. This distraction limit force, and the apparent ankle stiffness were patient-dependent, with ankle size and disease process playing major roles. Maximum forces developed were between 300 and 700 N (Table 1).Table 1: Intraoperatively and Radiographically Measured Parameters for Six SubjectsFig 3.: The graph shows force versus distraction length as measured by the instrumented distractor, for a typical subject. The ripples in the curves are associated with manual turns of the distractor. The initial trial’s distraction stiffness (slope of the force-versus-distraction curve) is clearly distinct from that in later (repeat) distraction trials. The shifting of the point of functional ligament laxity can be seen in later trials.Initial stiffness generally was lower than that seen in repeat trials (p = 0.028, paired t test). This can be seen clearly in Figure 3. Repeat trials showed an initial shift of the point of functional ligament laxity, defined as the distraction length where the ankle just begins to resist distraction and force begins to increase. The lateral incision, done to remove the syndesmosis (an integral step in the Agility Ankle technique), also seemed to have a reduction effect on the point of functional laxity and ankle stiffness. Mild relaxation in the distraction force occurred during the procedure, but the nominal value was maintained unless appreciable movement of foot or external fixator occurred. After reaching the distraction limit, the distractor was ″backed off″ typically 1 to 4 mm (except for Subject D, Table 1) before bone resection, to allow for subsequent insertion of the components. This backing-off substantially decreased the force measured in the instrumented distractor. The apparent nonlinear loss of distraction force is thought to be attributable to the effects of frictional resistance in the external fixator, as evidenced by large hysteresis curves shown in Figure 3. Ankle stiffness, defined as the slope of the distractor force versus displacement curve, was calculated for each ankle and is shown in Table 1. Radiographically assessed maximum joint opening, based on intraoperative fluoroscopic images, was available for four patients (right column, Table 1). Images recorded intraoperatively show the joint distraction achieved at four stages in the procedure for one typical patient (Fig 4).Fig 4.: A–D. Fluoroscopic images show an ankle under distraction, and with an implant. (A) The opening achieved at perceived maximum possible distraction can be seen. The substantial distraction is seen to involve relatively uniform opening of the joint on the medial and lateral sides. (B) The opening after distraction is backed off to increase bone resection and allow extra clearance for eventual implant insertion (9.7 mm). (C) A cutting guide is seen aligned with the ankle before resection (altered imaging angle of view). (D) The implant can be seen in place preserving the relative distance between tibia and talus after removal of the distraction force. The apparent opening between components is the PE bearing.Ligament tensioning, beyond the point of functional ligament laxity, seemed sufficient to stabilize the ankle in all patients except one. Lack of sufficient tensioning during surgery in Ankle D, as evidenced by the low value of distraction above functional laxity after back-off (Table 1), resulted in a visibly loose ankle when manipulated by hand postoperatively. In this case, previously planned ligament reconstruction tightened the joint substantially. The distraction stiffness of the ankle from a cadaver on which the simulated total ankle replacement was done was 80 N/mm. This was in the same range as the intraoperative measurements. No appreciable skew in joint opening was observed. The three intact ankles distracted without preconditioning in the materials testing machine had distraction stiffnesses of 100, 111, and 156 N/mm (mean, 122 N/mm). Stiffness curves were primarily linear, except for the initial toe-in region. Bench testing of the distractor and external fixator revealed that the internal frictional forces in the external fixator were approximately linearly proportional to the distraction force that was developed at the joint level. Likewise, the ratio of distractor torque to distractor force was approximately linear. calibration variability was seen in the between instrumented joint distraction is done using distraction length or joint opening as the for purposes of postoperative ligament tensioning, distraction force is a more appropriate Distraction opening could be used as a for force ankle stiffnesses were across The results of this however, show a large variability in ankle stiffnesses for patients having ankle replacement. distraction length were used, distraction force and ligament tension would by the same as the ankle stiffness, which could be than In either ligament tensioning or ligament injury easily could the external fixator had a substantial but effect on the distraction force as measured by the distractor unit. A previous study showed that internal appreciable resistance to in this external In perceived distraction force is by in the distractor unit and the external fixator, the relationship of torque versus force of distraction was shown to be nearly linear. This that the surgeon’s tactile sense of torque can be used to in the stiffness of the ankle, for a surgeon to achieve a specific distraction the relationship for of distractor and fixator unit used would have to be and the surgeon would have to the ability to apply an torque to the distractor unit. Unfortunately, the variability in the frictional of distractor observed in this study question the of distraction force by tactile at least using this distraction variability in the of distraction force achieved was observed between repeat distraction trials in the same variation was observed This variability could be attributable to the surgeon’s in applying more or less distraction force to the or intraoperatively apparent of each ankle. At least of the variability seems to be attributable to the difficulty of the of torque as it is being applied to turn the distractor unit, to the appropriate ligamentous tensioning. no evidence was of a in the force versus distraction curve, to account for the point of torque resistance used as a surgical for appropriate distraction. This of a torque and the observed variability in maximum forces measured to that no clearly point The perceived limit is a function of the increasing torque needed to the distractor internal and, is not appropriate for as a guide or That using this resistance as a surgical does not to of ligament tensioning either from patient to patient, or surgeon to surgeon. an force measurement during distraction may guide surgeons to achieve and more ligament tensioning during total ankle replacement. on force rather than distractor opening, more ligament tensioning could be while the of ligament measurement of forces applied to the distractor, as shown in this can the variability in of ankle distraction ankle stiffness is is for an additional The periarticular ligament after implant can be based on or measuring the in distraction or exists the distraction range of mm, then loss of 1 or 2 mm during the final process of implant insertion ligament tensioning by only substantial tensioning across the joint. relaxation in distraction force was observed with when distraction length was held Distraction beyond functional laxity was and from mm at the of laxity was in only one patient, Subject D, who had and tensioning of the ligaments during ankle replacement. These data the of the at least in the postoperative ligamentous tensioning is the implant is in The shifting of the point of functional laxity, seen in the data as a between initial and subsequent trials on the same ankle, most was attributable to of the soft and soft tissue the initial anterior generally done after the first distraction. After the initial distraction, subsequent trials showed total distraction measured from fluoroscopic images taken at the of resection can be with the made using the instrumented distractor. The actual distraction is measured from of the the instrumented distraction reported is measured from the point of functional ligament The between the two may be to ligament laxity by the loss of and the of the typical of degenerative joint The linear seen in study is with that observed for ankle ligaments by et ankle distraction stiffness values in study were lower than might be expected based on the results of et Stiffness values for the calcaneofibular ligament in that study N/mm and for the ligament of they N/mm N/mm). those values two to three the stiffness values that we intraoperatively. factors may have to this First, the relative of each ligament to the stiffness of the entire ankle would on each with to the of distraction. Second, the patients having total ankle in study had degenerative which a reduction in ankle stiffness. measured distraction stiffness is by pin and/or pin which of is and distraction trials that the range of intraoperative stiffnesses measured is ankle distraction of the specimens stiffness values just than those measured intraoperatively. Maximum force developed intraoperatively during distraction for total ankle replacement based on the reported force for the same two N for the calcaneofibular and N for the Even without in a from ligaments and soft of these two ligaments a range for maximum distraction forces to the maximum values measured in the In its the of distraction to ligament tensioning across the ankle is because of the distraction and of a that can be each distraction An distraction that could measure the actual force of ankle distraction may this a system would the surgeon the means to reproducible ankle ligament tensioning, thereby in the of ankle arthroplasty.