- Email: [email protected]
The Use of Navigation to Obtain Rectangular Flexion and Extension Gaps During Primary Total Knee Arthroplasty and Midterm Clinical Results
The Journal of Arthroplasty Vol. 26 No. 4 2011
The Use of Navigation to Obtain Rectangular Flexion and Extension Gaps During Primary Total Knee Arthroplasty and Midterm Clinical Results Jong-Keun Seon, MD, Eun-Kyoo Song, MD, PhD, Sang-Jin Park, MD, and Dam-Seon Lee, MD
Abstract: The authors evaluated 112 knees treated by total knee arthroplasty (TKA) using a navigation-assisted modified gap balancing technique. Initial mediolateral gap differences in extension and in 90° of flexion were measured after proximal tibia bone cutting. Final flexion and extension gaps were measured by checking distances under equal tension before prosthesis insertion. Amount of femoral bone cutting and external rotations of femoral components were found to depend on initial gaps. Patients with a final rectangular gap had greater knee flexion angles preoperatively and at 1 year after TKA. However, no differences were observed between the clinical and radiologic outcomes of knees with rectangular and nonrectangular gaps at 1 or 4 years after TKA. The study shows that the navigation-assisted modified gap balancing technique provides an effective means of achieving rectangular flexion and extension gaps during TKA. Keywords: gap technique, navigation, total knee arthroplasty. © 2011 Elsevier Inc. All rights reserved.
It is well known that successful total knee arthroplasty (TKA) depends on several factors, that is, proper patient selection, appropriate implant design, the correct surgical technique, and effective perioperative care. Accurate restoration of mechanical leg axis and proper knee stabilities in extension and flexion have been cited as the 2 most important aspects of successful TKA [1,2]. To achieve these objectives, balanced medial-lateral and flexion-extension gaps are important [3,4]. However, soft tissue balancing remains a difficult issue during TKA, because much depends on surgeon's “feel”[2,5]. The navigation systems increasingly used for TKA provide excellent restoration of the mechanical axis and precise component positioning, which improve the accuracy of the balancing procedure by providing more objective and quantitative measures of flexion and extension gaps [6,7]. Although soft tissue balancing is recognized an essential surgical intervention for improving TKA outFrom the Department of Orthopedic Surgery, Center for Joint Disease, Chonnam National University Hwasun Hospital, Jeonnam, South Korea. Submitted September 25, 2009; accepted April 25, 2010. No benefits or funds were received in support of this study. Reprint requests: Eun-Kyoo Song, MD, PhD, Department of Orthopaedic Surgery, Chonnam National University Hwasun Hospital, 160 Ilsimri, Hwasun-eup, Hwasun-gun, Jeonnam 519-809, South Korea. © 2011 Elsevier Inc. All rights reserved. 0883-5403/2604-0012$36.00/0 doi:10.1016/j.arth.2010.04.030
comes [3,4], few reports have described the relationship between flexion-extension gap differences and postoperative outcomes [8-10]. Unitt et al [11] reported shortterm outcomes for TKA after soft tissue release and balancing. In their study, gaps were measured using a balancing instrument, and the balancing of imbalanced knees was found to improve short-term knee outcomes. Our hypothesis was that that navigation provides a useful means of achieving rectangular flexion-extension gaps during TKA and that flexion-extension gap balancing affects clinical and radiologic outcomes. The objectives of this study were to investigate prospectively the efficacy of navigation in terms of obtaining rectangular flexion-extension gaps during TKA using the modified gap balancing technique and to evaluate whether flexion-extension gap differences affect the clinical and radiologic results of TKA.
Materials and Methods Eighty-eight patients (112 knees) who were performed TKA using the modified gap balancing technique for osteoarthritis and followed up for at least 4 years were included for this prospective study. There were 9 men and 79 women, with an average age of 67.6 years (range, 52-79 years). Mean follow-up duration was 52.8 months (range, 48–58 months). Orthopilot Version 4.0 or 4.2 (Aesculap, Tuttlingen, Germany) navigation systems were used in all cases. Ethical approval was granted for the study, and all patients who participated
582
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
provided informed consent. Total knee arthroplasties were performed using an e-motion prosthesis (B. Braun Aesculap). All procedures were performed using the navigation-assisted modified gap balancing technique by a single surgeon. Surgical Technique Patients were placed in a supine position, and surgery was carried out under a tourniquet. Using a midline skin incision and medial parapatellar approach, the knee joint was exposed. Two 4.5-mm bicortical screws were then inserted into the femur and tibia through separate stab incisions, and infrared light-emitting diode containing rigid bodies were attached to the screws. Kinematic registrations were performed sequentially at joint centers (hip, knee, and ankle). Anatomic landmarks were registered by hand using a pointer to define the joint line and the mechanical axis of the leg. Navigation provided a dynamic display by calculating the momentary axis relation between the mechanical tibial axis and the mechanical femoral axis. Stress testing was performed before the soft tissue balancing was made. The surgeon
583
stressed the knee and attempted to correct the deformity present to a reading of 0° on the computerized navigation system. Accordingly, sequential soft tissue release and osteophyte removal were performed until the varus or valgus deformity reduced to within 2° at 0° of knee extension. At this stage, the flexion gap was not a concern. After all osteophytes, except the posterior knee joint, had been removed from the tibia and femur, proximal tibial osteotomy was performed perpendicular to the mechanical axis in the coronal and sagittal planes using a cutting block positioned under navigation guidance. The new tibial height was then measured using a check-plate and inputted into the navigation system after tibial cutting. During osteotomy, the bony block at the tibial insertion site of the posterior cruciate ligament was preserved and the posterior cruciate ligament (PCL) was confirmed to be anatomically intact by inspecting and checking its tension by probing its fibers. At this stage, preliminary mediolateral gaps were measured at full extension and at 90° of flexion using a tensioning device (V-STAT tensor; Zimmer, Warsaw, Ind) and a special torque wrench set at 200 Nm. When
Fig. 1. After obtaining an acceptable extension gap by preliminary soft tissue release, we checked initial extension and flexion gaps before femoral planning, at 0° of extension (A) and at 90° of knee flexion (B).
584 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 initial extension gaps were tight on the medial side, the medial collateral ligament was released in a progressive manner until extension gap balance was achieved. Similarly, when initial extension gaps were tight on the lateral side, the lateral structure including posterolateral capsule and iliotibial band was progressively released. The computer showed that when the difference between medial and lateral gaps in extension was not more than 3 mm, a mechanical axis of 0° was achievable with a total deflection arc of 3 mm or less. We tried to balance gaps within 3 mm by stepwise soft tissue release in accord with preliminary gap differences in extension. After obtaining an acceptable extension gap with a mediolateral gap difference of 3 mm or less, we rechecked flexion and extension gaps, before femoral planning, at 0° of extension and at 90° of knee flexion, and refer to these as initial flexion-extension gaps (Fig. 1). Initial gap differences were classified as rectangular (group 1), tight
medial (group 2), and tight lateral (group 3) at 0° of extension and at 90° of knee flexion, respectively. A rectangular gap was defined as having a mediolateral gap difference of 3 mm or less, a tight medial gap was defined as one with a lateral gap of greater than 3 mm more than the corresponding medial gap, and a tight lateral gap was defined as one with a lateral gap of more than 3 mm less than the corresponding medial gap. To balance discrepancies between medial and lateral gaps at 90° of knee flexion, femoral rotation was adjusted to equalize flexion gaps. The OrthoPilot navigation system offers a femoral planning step that allows femoral component sizing and rotation to be simulated to achieve a balanced gap (Fig. 2). If less than 6° of external rotation (ER) of the femoral component was needed at the femoral planning step to achieve a balanced flexion gap, flexion gap asymmetry was corrected by rotating the femoral component externally to correct the deformity without any collateral ligament
Fig. 2. To balance discrepancies between medial and lateral gaps at 90° of knee flexion (A), ERs of femoral component was adjusted to equalize flexion gaps. Levels of distal and posterior femoral cuts and amounts of femoral component rotation were planned based on mediolateral flexion and extension gap gaps. At this stage, a posterior cutting depth of 2 mm more than the extension gap was planned (B). According to this, distal femoral cutting was done (C), and then, according to the planned ERs and size of femoral component, depths of posterior femoral condylar cutting were determined (D).
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
release, which would be obtained by increased posterior medial condyle cutting depth. We considered routine lateral reticular release when an ER of 6° or greater of the femoral component was needed to achieve flexion gap balance. In cases requiring more than 7° ER of the femoral component, we proceeded to soft tissue release to reduce medial flexion gap tightness and reduce the amount of ER of the femoral component required. In such cases, we gently released the anterior fiber of the superficial medial collateral ligament and rechecked flexion and extension gap balancing. Femoral ER was restricted to 7° or less to avoid deleterious patellar tracking effects, inline with personal experience. At this stage, a posterior cutting depth of 2 mm more than the extension gap was planned, and mediolateral flexion and extension gap differences with the patella in an everted position were checked before posterior osteophyte removal on the femoral side. Under these circumstances, the measured gap range would be overestimated by ∼2 mm. Based on personal experience, to obtain a finally balanced flexion and extension gaps, the initial flexion gap should be greater than 2 mm. Using
585
the femoral planning steps in navigation and the abovementioned method, balanced rectangular flexion-extension gaps were obtained. Distal and posterior femoral bone cutting was then performed using a navigation-assisted gap balancing technique. After cutting the posterior condyle, all posterior osteophytes were removed. The trial components were then inserted, and gap balance was checked manually applying varus and valgus force at full extension and at 90° of knee flexion. Flexion and extension gap balancing was checked by moving the limb from full extension to 90° of knee flexion. If the flexion gap was tighter than the extension gap at this step, PCL tension was checked at 90° of knee flexion using the probe after trial component insertion. When the PCL was tight, gradual PCL recession was performed from femoral insertion site. Flexion and extension gaps were generally balanced using this method. However, in some cases when the PCL was too tight to enable the flexion-extension gap to be balanced, the PCL was released entirely and a PCL sacrificing type of implant was used instead of a cruciate retaining implant.
Fig. 3. Finally, all trial components were removed and flexion-extension gaps were reassessed mediolaterally at full extension (A) and at 90° of knee flexion (B) before actual prosthesis insertion. Final flexion-extension gaps were also checked under equal tension mediolaterally in the same manner using a tensioning device (V-STAT tensor, Zimmer) and a special torque wrench set at 200 Nm.
586 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 Table 1. Clinical and Radiographic Outcomes of 112 Knees After Navigation-assisted TKA Using the Modified Gap Balancing Technique * Preoperative HSS score WOMAC score Extension lag (°) ROM (°) Mechanical axis (varus, °)
61.6 ± 77.2 ± 9.9 ± 115.8 ± 10.6 ±
Postoperative 1 Y
8.8 12.3 7.5 21.2 5.4
93.3 ± 4.4 31.3 ± 6.5 0.8 ± 2.2 124.0 ± 30.5 1.2 ± 1.9
Postoperative 4 Y 95.3 ± 29.1 ± 0.1 ± 132.5 ± 1.1 ±
5.7 8.0 0.7 7.4 1.8
*Significantly improved clinical results were obtained at 1 year postoperatively for all parameters (P = .000). Statistical significance was accepted for P value of b.05, as determined by the paired t test. Only WOMAC scores improved significantly between 1 and 4 years postoperatively (P = .001).
Finally, all trial components were removed, and flexion-extension gaps were reassessed mediolaterally at full extension and at 90° of knee flexion before actual prosthesis insertion (Fig. 3). Final flexion-extension gaps were also checked under equal tension mediolaterally in the same manner using a tensioning device (V-STAT tensor, Zimmer) and a special torque wrench set at 200 Nm. Final gap differences were classified based on final flexion and extension gap differences; rectangular (group 1), tight medial (group 2), and tight lateral (group 3) at 0° of extension and at 90° of knee flexion, using the gap classification criteria described above. Data about levels of distal and posterior femoral cuts and amounts of femoral component rotations were obtained using the navigation monitor. If mediolateral rectangular gaps were obtained at full extension and at 90° of knee flexion, the knee concerned was allocated to group A, and if mediolateral gaps were not rectangular at full extension or at 90° of knee flexion, the knee concerned was allocated to group B. These 2 groups were compared with respect to intraoperative data and clinical and radiologic outcomes (Table 4). Clinical outcomes were assessed using Hospital for Special Surgery (HSS) scores and Western Ontario MacMaster (WOMAC) scores, which were checked preoperatively and at 1 and 4 years postoperatively, respectively. Mechanical limb alignment was checked using a standing radiograph of the entire lower extremity. Passive maximum knee range of motion (ROM) was measured using a goniometer. All assessments were performed by 1 author, not directly involved
in the surgical procedures, before and at 1 and 4 years after surgery. Statistical Analysis Collection and analysis of data were performed using SPSS version 12 (SPSS Inc, Chicago, Ill). Outcome measures were assessed in 3 different ways. First, clinical outcomes were assessed preoperatively and at 1 and 4 years postoperatively, and improvements were compared using the paired t test. Second, knees were analyzed according to initial mediolateral gap differences and allocated to the rectangular (group 1), the tight medial (group 2), and the tight lateral gaps (group 3) at full extension and at 90° of knee flexion, respectively. Analysis of variance (ANOVA) was used to assess differences between these 3 groups, and the post hoc Bonferroni test was applied when significance differences were found. A third analysis was performed to examine the extents of final flexion and extension mediolateral gap differences, which were classified as final rectangular (group A) and final nonrectangular (group B). The independent t test was used to compare groups A and B. Statistical significance was accepted for P values ≤.05.
Results Considering all patients (112 knees), mean (SD) HSS scores improved from a preoperative score of 61.6 (8.8) to 93.3 (4.4) at 1 year postoperatively and to 95.3 (5.7) at 4 years postoperatively. Mean (SD) WOMAC scores improved from 77.2 (12.3) preoperative to 31.3 (6.5) at 1 year postoperatively and to 29.1 (8.0) at 4 years
Table 2. Comparisons of Level of Femoral Bone Cutting According to Initial Mediolateral Gaps at 0° of Knee Extension
ER of femoral component (°) Medial distal cutting (mm) Lateral distal cutting (mm) Medial posterior cutting (mm) Lateral posterior cutting (mm)
Group 1
Group 2 *
Group 3
Rectangular Gap (n = 73)
Tight Medial Gap (n = 34)
Tight Lateral Gap (n = 5)
P-Value (ANOVA †)
4.5 ± 1.2 7.8 ± 2.6 9.3 ± 1.8 8.5 ± 4.9 7.1 ± 1.7
.016 † .000 † .756 .000 † .369
2.8 ± 2.1 8.7 ± 2.8 8.9 ± 3.0 11.7 ± 3.1 8.8 ± 2.8
4.0 ± 2.4 11.9 ± 3.4 8.5 ± 3.2 14.0 ± 2.6 9.1 ± 3.0
*Group 2 required more ER of femoral component than did group 1(P = .032), more medial distal cutting than did groups 1 (P = .000) and 3 (P = .020), and more medial posterior cutting than did groups 1 (P = .006) and 3 (P = .001). †The 3 groups were compared using ANOVA, and statistical significance was accepted for P values less than .05. Intergroup differences were assessed using the post hoc Bonferroni test.
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
587
Table 3. Comparisons of Levels of Intraoperative Femoral Bone Cutting According to Initial Mediolateral Gaps in 90° of Knee Flexion
ER of femoral component (°) Medial distal cutting (mm) Lateral distal cutting (mm) Medial posterior cutting (mm) Lateral posterior cutting (mm)
Group 1
Group 2
Group 3
Rectangular Gap (n = 64)
Tight Medial Gap (n = 41)
Tight Lateral Gap (n = 7)
2.2 ± 9.4 ± 9.2 ± 12.0 ± 9.6 ±
1.8 3.1 3.1 2.9 2.8
4.7 ± 10.0 ± 8.3 ± 13.3 ± 7.6 ±
1.9 † 3.8 3.0 3.2 2.7 †
3.4 ± 8.9 ± 9.1 ± 8.1 ± 8.8 ±
2.4 1.9 2.1 3.9 ‡ 1.8
P (ANOVA ) .000 * .637 .437 .000 * .005 *
*The 3 groups were compared using ANOVA, and statistical significance was accepted for P values less than .05. Intergroup differences were assessed using the post hoc Bonferroni test. †Group 2 required more ER of the femoral component (P = .000) and less lateral posterior femoral bone cutting (P = .004) than did group 1. ‡In group 3, medial posterior cutting amounts were smaller than in groups 1 and 2.
postoperatively. The mean (SD) knee ROMs improved from 115.8° (21.2°) preoperatively to 124.0° (30.5°) at 1 year and to 132.5° (7.4°) at 4 years postoperatively, and mean (SD) alignment of the lower extremity mechanical axis improved from a varus value of 10.6° (5.4°) preoperative to 1.2° of varus (1.9°) at 1 year and to 1.1° of varus (SD 1.8°) at 4 years postoperatively. Significant improvements in clinical outcomes were obtained for all parameters at 1 year postoperatively. With regard to outcomes at 4 years postoperatively, only WOMAC scores improved significantly as compared with results at 1 year postoperatively (P = .001; Table 1). Comparisons of Clinical Results and Intraoperative Parameters of Femoral Bone Cutting in Relation to Initial Mediolateral Gap Differences Patients were grouped according to initial mediolateral gap differences. At zero degrees of knee extension, there were 73 knees with a rectangular gap (group 1), 34 knees with a tight medial gap (group 2), and 5 knees
with a tight lateral gap (group 3). Radiographic evaluations revealed significant differences in lower extremity alignment according to initial gap differences. In particular, more varus alignment was observed in group 2 (mechanical axis, varus 2.22°) than in the other 2 groups at 1 year postoperatively (group 1, varus 3.24°; group 3, varus 0.04°; P = .001). However, no significant difference was observed between these 3 groups in terms of preoperative and postoperative mean ROM and HSS or WOMAC scores. Amounts of ER of the femoral component and of medial distal cutting and medial posterior cutting of femur were significantly different among the initial gap groups at zero degrees of knee extension. Intergroup differences were assessed using the post hoc Bonferroni test. Group 2 showed more ER of the femoral component than did group 1 (P = .032), more medial distal cutting than did groups 1 (P = .000) and 3 (P = .020), and more medial posterior cutting than did groups 1 (P = .006) or 3 (P = .001; Table 2). According to initial
Table 4. Comparison of Clinical and Radiologic Results According to Final Mediolateral Gap Differences
HSS score Preoperative Postoperative 1 y Postoperative 4 y WOMAC score Preoperative Postoperative 1 y Postoperative 4 y Extension lag (°) Preoperative Postoperative 1 y Postoperative 4 y Further flexion (°) Preoperative Postoperative 1 y Postoperative 4 y Mechanical axis (varus, °) Preoperative Postoperative 1 y Postoperative 4 y
Group A
Group B
Rectangular Gap in Both Angles (n = 105)
Nonrectangular Gap in Any Angle (n = 7)
P (t test) *
61.7 ± 8.8 93.4 ± 4.3 94.5 ± 8.1
59.7 ± 9.8 91.8 ± 5.1 90.8 ± 5.4
.591 .406 .319
77.2 ± 10.9 31.3 ± 5.9 28.5 ± 6.3
76.7 ± 16.8 32.5 ± 7.4 31.3 ± 2.3
.944 .639 .327
9.5 ± 7.4 0.8 ± 2.2 0.2 ± 1.1
16.4 ± 6.3 1.7 ± 2.6 2.0 ± 4.5
.018 .337 .414
130.5 ± 14.9 132.6 ± 6.0 132.5 ± 6.8
100 ± 32.7 119.2 ± 10.7 135.0 ± 5.0
.048 .027 .424
10.7 ± 5.2 1.2 ± 1.8 1.0 ± 1.8
9.8 ± 8.0 0.8 ± 3.5 2.4 ± 1.6
.795 .776 .463
*Statistical analysis was performed using the independent t test, and significance was accepted for P values less than .05.
588 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 mediolateral gap differences at 90° of knee flexion, ERs of the femoral component and medial and lateral posterior femoral cutting amounts were significantly different between the initial flexion gap groups. Group 2 required more ER of the femoral component (P = .000) and less lateral posterior femoral bone cutting (P = .004) than did group 1, and groups 1 and 2 required more medial posterior cutting than did group 3 (Table 3). Comparison of Clinical and Radiologic Results in Relation to Final Mediolateral Gap Differences In most cases, we obtained a balanced extension gap by precise and stepwise release in extension, and most flexion gaps were balanced by adjusting initial posterior femoral bony cutting and ligament release. Patients were allocated to the final rectangular or nonrectangular gap groups according to final mediolateral gap difference at zero degrees of knee extension and at 90° of knee flexion. There were 105 knees with rectangular gaps (group A), in which gaps were balanced at full extension and at 90° of knee flexion, and 7 knees with nonrectangular gaps (group B). However, clinical HSS and WOMAC scores and preoperative and postoperative mechanical axes of the lower extremities were no different in these 2 groups. Group A showed a significantly better flexion range preoperatively and at 1 year postoperatively (Table 4). With regard to intraoperative parameters, no differences were found between these 2 groups with respect to femoral cutting levels or ER of the femoral component. Complications Two patients experienced a superficial wound infection but responded to intravenous antibiotics. There were 3 cases of symptomatic deep-vein thrombosis, but all cured spontaneously. No knee showed radiolucency, osteolysis, subsidence, or loosening during follow-up.
Discussion Two surgical techniques can be used to establish ligament balance, namely, the classic flexion-extension gap balancing technique and the measured resection technique. According to the measured resection technique, the surgeon removes from the distal femur and proximal tibia the amount of bone being replaced by the prosthetic component. Therefore, distal femoral resection and posterior femoral condylar resection thicknesses are equal to those of the metallic femoral component, whereas proximal tibial resection approximates to the thickness of the tibial component. Soft tissue is balanced to obtain stable and symmetrical flexion-extension gaps, but this technique may have accompanying problems in imbalanced patients. To resolve an imbalanced gap, multiple additional bone cuts or soft tissue releases are needed after complete tibial and femoral bone cuts have been made. However,
redo multiple bone cuts or soft tissue releases could create more imbalance. Conversely, the classic flexion-extension gap balancing technique is based on the appropriate sizing of the femur with respect to the anteroposterior dimension. Posterior femoral condylar resection, therefore, varies as determined by this sizing of the femur. Proximal tibial resection is performed to remove approximately 2 mm of bone from the more deficient tibial plateau. These 2 resections determine the dimension of the flexion gap, and the extension gap is created to match these dimensions. If appropriate posterior soft tissue release has not been performed, this technique is more susceptible to erroneous distal femoral resection than the measured resection technique, and therefore, a number of surgeons who have traditionally used the classic flexion-extension gap balancing technique now favor a combination technique, referred to as the modified gap technique, which attempts to blend the measured resection and the classic flexionextension gap balancing techniques. In our study, the tibia-first technique was used at 0° in both the coronal and sagittal planes, to prepare extension gaps before posterior femoral condyle cutting (flexion gap preparation). Cutting the femur first can deprive the surgeon of the many choices that can be readily calculated using the computer, such as balancing of the knee using adjustable femoral cuts. For example, if a surgeon tries to prepare the flexion gap first, the shape and size of the flexion gap should be determined before axial alignment is adjusted. However, we consider this to be very difficult, and furthermore, this difficulty is compounded when the knee has an advanced deformity. Accordingly, we advocate an alternative method, namely, that the rectangular extension gap be prepared first and that this be followed by an equal flexion gap. Navigation systems now provide femoral planning based on initial flexion and extension gap measurements. Based on gap differences, distal femur cutting and posterior condylar cutting depth can be planned and femoral component size and rotation adjustments can be simulated to achieve flexion and extension gap balance. Furthermore, final extension gaps can be adjusted during navigation-assisted TKA by modifying the distal femur cutting depth and flexion gap configuration. During this stage, we did not adjust the coronal cutting angle to obtain cutting perpendicular to the femoral mechanical axis, and thus, soft tissue release is most important task in terms of providing a neutral coronal limb axis. As can be done during the modified gap technique, the ER of the femoral component can be adjusted to obtain a balanced flexion gap, mediolaterally. However, in cases with excessive ER of the femoral component, the specific portion of soft tissue responsible for the tight flexion gap must be released to avoid patellofemoral problem.
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
Until recently, intraoperative gap balancing measurement has depended on subjective assessments made by operating surgeons. However, in the present study, we used a calibrated torque wrench and tensor system as a means of obtaining reproducible measures of gap differences, and we believe that this method could be used to help surgeons determine optimal balance in individual knees earlier during operations. Unitt et al [11] measured flexion-extension gaps in 218 TKAs using the measured resection technique before and after soft tissue release and defined a nonbalanced gap as one with more than 3 mm of laxity. Based on their results, balanced flexion and extension gaps during TKA were achieved using the measured resection technique in 175 knees (80.3%). In the present study, we obtained a final rectangular gap in 105 knees (94%) using the navigation-assisted gap balancing technique. When we compared the results of the present study with those of the study of Unitt et al, we found that TKA using the navigation-assisted gap balancing technique produced better balanced flexion and extension gaps than TKA using the measured resection technique. Relatively few studies have addressed the influence of soft-tissue balancing on TKA outcomes [11-13]. Attfield et al [12] reported that knees balanced in both full extension and in flexion showed improved proprioception postoperatively, and in the present study, significant improvements in clinical outcomes were observed at 1 year postoperatively after TKA. However, only WOMAC scores showed a significant improvement at 4 years postoperatively versus 1 year postoperatively. Lower extremity alignments and good clinical outcomes were also maintained at 4 years after TKA. However, no association was found between final gap differences and preoperative to postoperative differences in terms of mechanical axis alignments or clinical outcomes, such as HSS and WOMAC scores. However, although we tried to achieve a rectangular gap, we failed to do so in 7 patients, which were particularly true for flexion gaps, although no patient required a constrained prosthesis. The minimum gap difference in the nonrectangular gap group was 5 mm, and even though a rectangular final gaps were not achieved when measurements were conducted using a tensioner, knees were within the acceptable range of instability clinically by manual testing, and thus, results in these cases were accepted. This may explain the lack of a difference between the outcomes of balanced and imbalanced knees. However, patients with greater preoperative flexion ranges were found to be more likely to have a rectangular final mediolateral gap than a nonrectangular gap. Furthermore, we believe that it was easier to achieve rectangular gaps in knees with less deformity, although by using the navigationassisted modified gap balancing technique, we were able to obtain a rectangular gap at similar rates in
589
coronally deformed and less deformed knees. With regard to ROM, those with well-flexed knees tended to have greater flexion gaps during surgery, which resulted in greater flex postoperatively if a wellbalanced gap was obtained during index surgery. Navigation systems are being increasingly used for TKA and now provide excellent mechanical axis restoration and precise component positioning. In addition, these improvements increase the accuracy of the balancing procedure by providing more objective and quantitative measures of flexion and extension gaps [6]. In the present study, we defined a rectangular gap as one with a flexion-extension difference of 3 mm or less, a tight flexion gap as one with an extension gap of 3 mm more than the corresponding flexion gap, and a tight extension gap as one with a flexion gap of 3 mm more than the corresponding extension gap. To date, no study has been undertaken to determine the relation between the amount of femoral bone cut and initial and final gap differences. In the present study, we analyzed relations between initial gap differences and the amounts of femoral bone cut and the ERs of femoral components. In the tight medial gap group at 0° of knee extension, more medial side distal and posterior femoral bone cutting were needed to balance final gaps. This suggests that degrees of femoral bone cutting and ER can be predicted from initial gap differences after proximal tibial bone cutting and preliminary soft tissue release. In the present study, medial posterior femoral bone cutting amounts were found to be related to initial gap difference at 90° of flexion and at 0° of knee extension. Furthermore, the ERs of femoral components were found to be more related with initial gap differences at 90° of flexion than in extension. In addition, the tight medial gap group at 90° of flexion needed a mean 2.5° of ER more than the balanced group. In the rectangular gap group, amounts of bone cutting and the ER of femoral implants during the gap balancing technique were similar to those of the measured resection technique. In terms of the strength of this study, we provided the surgeon with information regarding the amount of bone cutting required based on initial gaps, and we attempted to create a rectangular space using same tensions during gap balancing. However, the limitations of this study are that it does not contain a control group or a nonnavigated group and that the final nonrectangular gap group was too small to provide comparative data. Further study is needed to compare the clinical and radiologic outcomes of patients with balanced and imbalanced gaps. In conclusion, amount of intraoperative femoral bone cutting and the ERs of femoral components were found to depend on initial gap differences. However, no differences in clinical or radiologic outcomes were observed between the final rectangular and nonrectangular groups at 1 or 4 years after TKA. Furthermore,
590 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 navigation was found to provide information about the amount of bone cutting required based on initial gap differences. In addition, we found that navigation provides an effective means of achieving rectangular flexion-extension gaps during TKA performed using the navigation-assisted modified gap balancing technique.
References 1. Freeman MA, Todd RC, Bamert P, et al. ICLH Arthroplasty of the knee: 1968-1977. J bone Joint Surg Br 1978;660-B: 339. 2. Insall JN, Binazzi R, Soudry M, et al. Total knee arthroplasty. Clin Orthop 1985;192:13. 3. Fehring TK, Odum S, Griffin WL, et al. Early failure in total knee arthroplasty. Clin Orthop Relat Res 2001;392:315. 4. Sharkey PF, Hozack WJ, Rothman RH, et al. Insall award paper. Why are total knee arthroplasties failing today? Clin Orthop Relat Res 2002;404:7. 5. Dorr LD, Boiardo RA. Technical considerations in total knee arthroplasty. Clin Orthop 1986;205:5. 6. Clemens U, Miehlke RK. Advanced navigation planning including soft tissue management. Orthopedics 2005;28 (10 Suppl):s1259.
7. Matsumoto T, Muratsu H, Tsumura N, et al. Soft tissue balance measurement in posterior-stabilized total knee arthroplasty with a navigation system. J Arthroplasty 2009;24:358. 8. Han SB, Nha KW, Yoon JR, et al. The reliability of navigation-guided gap technique in total knee arthroplasty. Orthopedics 2008;31(10 Supp1) pii: orthosupersite. com/view.asp?rID=35542. 9. Higuchi H, Hatayama K, Shimizu M, et al. Relationship between joint gap difference and range of motion in total knee arthroplasty: a prospective randomized study between different platforms. Int Orthop 2009;33:997. 10. Romero J, Stahelin T, Binkert C, et al. The clinical consequences of flexion gap asymmetry in total knee arthroplasty. J Arthroplasty 2007;22:235. 11. Unitt L, Sambatakakis A, Johnstone D, et al. Short-term outcome in total knee replacement after soft-tissue release and balancing. J Bone Joint Surg Br 2008;90-B:159. 12. Attfield SF, Wilton TJ, Pratt DJ, et al. Soft tissue balance and recovery of propioception after total knee replacement. J Bone Joint Surg Br 1996;78-B:540. 13. Mihalko WM, Whiteside LA, Krackow KA. Comparison of ligament balancing techniques during total knee arthroplasty. J Bone Joint Surg Am 2003;85-A(Suppl 4):132.
The Use of Navigation to Obtain Rectangular Flexion and Extension Gaps During Primary Total Knee Arthroplasty and Midterm Clinical Results Jong-Keun Seon, MD, Eun-Kyoo Song, MD, PhD, Sang-Jin Park, MD, and Dam-Seon Lee, MD
Abstract: The authors evaluated 112 knees treated by total knee arthroplasty (TKA) using a navigation-assisted modified gap balancing technique. Initial mediolateral gap differences in extension and in 90° of flexion were measured after proximal tibia bone cutting. Final flexion and extension gaps were measured by checking distances under equal tension before prosthesis insertion. Amount of femoral bone cutting and external rotations of femoral components were found to depend on initial gaps. Patients with a final rectangular gap had greater knee flexion angles preoperatively and at 1 year after TKA. However, no differences were observed between the clinical and radiologic outcomes of knees with rectangular and nonrectangular gaps at 1 or 4 years after TKA. The study shows that the navigation-assisted modified gap balancing technique provides an effective means of achieving rectangular flexion and extension gaps during TKA. Keywords: gap technique, navigation, total knee arthroplasty. © 2011 Elsevier Inc. All rights reserved.
It is well known that successful total knee arthroplasty (TKA) depends on several factors, that is, proper patient selection, appropriate implant design, the correct surgical technique, and effective perioperative care. Accurate restoration of mechanical leg axis and proper knee stabilities in extension and flexion have been cited as the 2 most important aspects of successful TKA [1,2]. To achieve these objectives, balanced medial-lateral and flexion-extension gaps are important [3,4]. However, soft tissue balancing remains a difficult issue during TKA, because much depends on surgeon's “feel”[2,5]. The navigation systems increasingly used for TKA provide excellent restoration of the mechanical axis and precise component positioning, which improve the accuracy of the balancing procedure by providing more objective and quantitative measures of flexion and extension gaps [6,7]. Although soft tissue balancing is recognized an essential surgical intervention for improving TKA outFrom the Department of Orthopedic Surgery, Center for Joint Disease, Chonnam National University Hwasun Hospital, Jeonnam, South Korea. Submitted September 25, 2009; accepted April 25, 2010. No benefits or funds were received in support of this study. Reprint requests: Eun-Kyoo Song, MD, PhD, Department of Orthopaedic Surgery, Chonnam National University Hwasun Hospital, 160 Ilsimri, Hwasun-eup, Hwasun-gun, Jeonnam 519-809, South Korea. © 2011 Elsevier Inc. All rights reserved. 0883-5403/2604-0012$36.00/0 doi:10.1016/j.arth.2010.04.030
comes [3,4], few reports have described the relationship between flexion-extension gap differences and postoperative outcomes [8-10]. Unitt et al [11] reported shortterm outcomes for TKA after soft tissue release and balancing. In their study, gaps were measured using a balancing instrument, and the balancing of imbalanced knees was found to improve short-term knee outcomes. Our hypothesis was that that navigation provides a useful means of achieving rectangular flexion-extension gaps during TKA and that flexion-extension gap balancing affects clinical and radiologic outcomes. The objectives of this study were to investigate prospectively the efficacy of navigation in terms of obtaining rectangular flexion-extension gaps during TKA using the modified gap balancing technique and to evaluate whether flexion-extension gap differences affect the clinical and radiologic results of TKA.
Materials and Methods Eighty-eight patients (112 knees) who were performed TKA using the modified gap balancing technique for osteoarthritis and followed up for at least 4 years were included for this prospective study. There were 9 men and 79 women, with an average age of 67.6 years (range, 52-79 years). Mean follow-up duration was 52.8 months (range, 48–58 months). Orthopilot Version 4.0 or 4.2 (Aesculap, Tuttlingen, Germany) navigation systems were used in all cases. Ethical approval was granted for the study, and all patients who participated
582
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
provided informed consent. Total knee arthroplasties were performed using an e-motion prosthesis (B. Braun Aesculap). All procedures were performed using the navigation-assisted modified gap balancing technique by a single surgeon. Surgical Technique Patients were placed in a supine position, and surgery was carried out under a tourniquet. Using a midline skin incision and medial parapatellar approach, the knee joint was exposed. Two 4.5-mm bicortical screws were then inserted into the femur and tibia through separate stab incisions, and infrared light-emitting diode containing rigid bodies were attached to the screws. Kinematic registrations were performed sequentially at joint centers (hip, knee, and ankle). Anatomic landmarks were registered by hand using a pointer to define the joint line and the mechanical axis of the leg. Navigation provided a dynamic display by calculating the momentary axis relation between the mechanical tibial axis and the mechanical femoral axis. Stress testing was performed before the soft tissue balancing was made. The surgeon
583
stressed the knee and attempted to correct the deformity present to a reading of 0° on the computerized navigation system. Accordingly, sequential soft tissue release and osteophyte removal were performed until the varus or valgus deformity reduced to within 2° at 0° of knee extension. At this stage, the flexion gap was not a concern. After all osteophytes, except the posterior knee joint, had been removed from the tibia and femur, proximal tibial osteotomy was performed perpendicular to the mechanical axis in the coronal and sagittal planes using a cutting block positioned under navigation guidance. The new tibial height was then measured using a check-plate and inputted into the navigation system after tibial cutting. During osteotomy, the bony block at the tibial insertion site of the posterior cruciate ligament was preserved and the posterior cruciate ligament (PCL) was confirmed to be anatomically intact by inspecting and checking its tension by probing its fibers. At this stage, preliminary mediolateral gaps were measured at full extension and at 90° of flexion using a tensioning device (V-STAT tensor; Zimmer, Warsaw, Ind) and a special torque wrench set at 200 Nm. When
Fig. 1. After obtaining an acceptable extension gap by preliminary soft tissue release, we checked initial extension and flexion gaps before femoral planning, at 0° of extension (A) and at 90° of knee flexion (B).
584 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 initial extension gaps were tight on the medial side, the medial collateral ligament was released in a progressive manner until extension gap balance was achieved. Similarly, when initial extension gaps were tight on the lateral side, the lateral structure including posterolateral capsule and iliotibial band was progressively released. The computer showed that when the difference between medial and lateral gaps in extension was not more than 3 mm, a mechanical axis of 0° was achievable with a total deflection arc of 3 mm or less. We tried to balance gaps within 3 mm by stepwise soft tissue release in accord with preliminary gap differences in extension. After obtaining an acceptable extension gap with a mediolateral gap difference of 3 mm or less, we rechecked flexion and extension gaps, before femoral planning, at 0° of extension and at 90° of knee flexion, and refer to these as initial flexion-extension gaps (Fig. 1). Initial gap differences were classified as rectangular (group 1), tight
medial (group 2), and tight lateral (group 3) at 0° of extension and at 90° of knee flexion, respectively. A rectangular gap was defined as having a mediolateral gap difference of 3 mm or less, a tight medial gap was defined as one with a lateral gap of greater than 3 mm more than the corresponding medial gap, and a tight lateral gap was defined as one with a lateral gap of more than 3 mm less than the corresponding medial gap. To balance discrepancies between medial and lateral gaps at 90° of knee flexion, femoral rotation was adjusted to equalize flexion gaps. The OrthoPilot navigation system offers a femoral planning step that allows femoral component sizing and rotation to be simulated to achieve a balanced gap (Fig. 2). If less than 6° of external rotation (ER) of the femoral component was needed at the femoral planning step to achieve a balanced flexion gap, flexion gap asymmetry was corrected by rotating the femoral component externally to correct the deformity without any collateral ligament
Fig. 2. To balance discrepancies between medial and lateral gaps at 90° of knee flexion (A), ERs of femoral component was adjusted to equalize flexion gaps. Levels of distal and posterior femoral cuts and amounts of femoral component rotation were planned based on mediolateral flexion and extension gap gaps. At this stage, a posterior cutting depth of 2 mm more than the extension gap was planned (B). According to this, distal femoral cutting was done (C), and then, according to the planned ERs and size of femoral component, depths of posterior femoral condylar cutting were determined (D).
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
release, which would be obtained by increased posterior medial condyle cutting depth. We considered routine lateral reticular release when an ER of 6° or greater of the femoral component was needed to achieve flexion gap balance. In cases requiring more than 7° ER of the femoral component, we proceeded to soft tissue release to reduce medial flexion gap tightness and reduce the amount of ER of the femoral component required. In such cases, we gently released the anterior fiber of the superficial medial collateral ligament and rechecked flexion and extension gap balancing. Femoral ER was restricted to 7° or less to avoid deleterious patellar tracking effects, inline with personal experience. At this stage, a posterior cutting depth of 2 mm more than the extension gap was planned, and mediolateral flexion and extension gap differences with the patella in an everted position were checked before posterior osteophyte removal on the femoral side. Under these circumstances, the measured gap range would be overestimated by ∼2 mm. Based on personal experience, to obtain a finally balanced flexion and extension gaps, the initial flexion gap should be greater than 2 mm. Using
585
the femoral planning steps in navigation and the abovementioned method, balanced rectangular flexion-extension gaps were obtained. Distal and posterior femoral bone cutting was then performed using a navigation-assisted gap balancing technique. After cutting the posterior condyle, all posterior osteophytes were removed. The trial components were then inserted, and gap balance was checked manually applying varus and valgus force at full extension and at 90° of knee flexion. Flexion and extension gap balancing was checked by moving the limb from full extension to 90° of knee flexion. If the flexion gap was tighter than the extension gap at this step, PCL tension was checked at 90° of knee flexion using the probe after trial component insertion. When the PCL was tight, gradual PCL recession was performed from femoral insertion site. Flexion and extension gaps were generally balanced using this method. However, in some cases when the PCL was too tight to enable the flexion-extension gap to be balanced, the PCL was released entirely and a PCL sacrificing type of implant was used instead of a cruciate retaining implant.
Fig. 3. Finally, all trial components were removed and flexion-extension gaps were reassessed mediolaterally at full extension (A) and at 90° of knee flexion (B) before actual prosthesis insertion. Final flexion-extension gaps were also checked under equal tension mediolaterally in the same manner using a tensioning device (V-STAT tensor, Zimmer) and a special torque wrench set at 200 Nm.
586 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 Table 1. Clinical and Radiographic Outcomes of 112 Knees After Navigation-assisted TKA Using the Modified Gap Balancing Technique * Preoperative HSS score WOMAC score Extension lag (°) ROM (°) Mechanical axis (varus, °)
61.6 ± 77.2 ± 9.9 ± 115.8 ± 10.6 ±
Postoperative 1 Y
8.8 12.3 7.5 21.2 5.4
93.3 ± 4.4 31.3 ± 6.5 0.8 ± 2.2 124.0 ± 30.5 1.2 ± 1.9
Postoperative 4 Y 95.3 ± 29.1 ± 0.1 ± 132.5 ± 1.1 ±
5.7 8.0 0.7 7.4 1.8
*Significantly improved clinical results were obtained at 1 year postoperatively for all parameters (P = .000). Statistical significance was accepted for P value of b.05, as determined by the paired t test. Only WOMAC scores improved significantly between 1 and 4 years postoperatively (P = .001).
Finally, all trial components were removed, and flexion-extension gaps were reassessed mediolaterally at full extension and at 90° of knee flexion before actual prosthesis insertion (Fig. 3). Final flexion-extension gaps were also checked under equal tension mediolaterally in the same manner using a tensioning device (V-STAT tensor, Zimmer) and a special torque wrench set at 200 Nm. Final gap differences were classified based on final flexion and extension gap differences; rectangular (group 1), tight medial (group 2), and tight lateral (group 3) at 0° of extension and at 90° of knee flexion, using the gap classification criteria described above. Data about levels of distal and posterior femoral cuts and amounts of femoral component rotations were obtained using the navigation monitor. If mediolateral rectangular gaps were obtained at full extension and at 90° of knee flexion, the knee concerned was allocated to group A, and if mediolateral gaps were not rectangular at full extension or at 90° of knee flexion, the knee concerned was allocated to group B. These 2 groups were compared with respect to intraoperative data and clinical and radiologic outcomes (Table 4). Clinical outcomes were assessed using Hospital for Special Surgery (HSS) scores and Western Ontario MacMaster (WOMAC) scores, which were checked preoperatively and at 1 and 4 years postoperatively, respectively. Mechanical limb alignment was checked using a standing radiograph of the entire lower extremity. Passive maximum knee range of motion (ROM) was measured using a goniometer. All assessments were performed by 1 author, not directly involved
in the surgical procedures, before and at 1 and 4 years after surgery. Statistical Analysis Collection and analysis of data were performed using SPSS version 12 (SPSS Inc, Chicago, Ill). Outcome measures were assessed in 3 different ways. First, clinical outcomes were assessed preoperatively and at 1 and 4 years postoperatively, and improvements were compared using the paired t test. Second, knees were analyzed according to initial mediolateral gap differences and allocated to the rectangular (group 1), the tight medial (group 2), and the tight lateral gaps (group 3) at full extension and at 90° of knee flexion, respectively. Analysis of variance (ANOVA) was used to assess differences between these 3 groups, and the post hoc Bonferroni test was applied when significance differences were found. A third analysis was performed to examine the extents of final flexion and extension mediolateral gap differences, which were classified as final rectangular (group A) and final nonrectangular (group B). The independent t test was used to compare groups A and B. Statistical significance was accepted for P values ≤.05.
Results Considering all patients (112 knees), mean (SD) HSS scores improved from a preoperative score of 61.6 (8.8) to 93.3 (4.4) at 1 year postoperatively and to 95.3 (5.7) at 4 years postoperatively. Mean (SD) WOMAC scores improved from 77.2 (12.3) preoperative to 31.3 (6.5) at 1 year postoperatively and to 29.1 (8.0) at 4 years
Table 2. Comparisons of Level of Femoral Bone Cutting According to Initial Mediolateral Gaps at 0° of Knee Extension
ER of femoral component (°) Medial distal cutting (mm) Lateral distal cutting (mm) Medial posterior cutting (mm) Lateral posterior cutting (mm)
Group 1
Group 2 *
Group 3
Rectangular Gap (n = 73)
Tight Medial Gap (n = 34)
Tight Lateral Gap (n = 5)
P-Value (ANOVA †)
4.5 ± 1.2 7.8 ± 2.6 9.3 ± 1.8 8.5 ± 4.9 7.1 ± 1.7
.016 † .000 † .756 .000 † .369
2.8 ± 2.1 8.7 ± 2.8 8.9 ± 3.0 11.7 ± 3.1 8.8 ± 2.8
4.0 ± 2.4 11.9 ± 3.4 8.5 ± 3.2 14.0 ± 2.6 9.1 ± 3.0
*Group 2 required more ER of femoral component than did group 1(P = .032), more medial distal cutting than did groups 1 (P = .000) and 3 (P = .020), and more medial posterior cutting than did groups 1 (P = .006) and 3 (P = .001). †The 3 groups were compared using ANOVA, and statistical significance was accepted for P values less than .05. Intergroup differences were assessed using the post hoc Bonferroni test.
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
587
Table 3. Comparisons of Levels of Intraoperative Femoral Bone Cutting According to Initial Mediolateral Gaps in 90° of Knee Flexion
ER of femoral component (°) Medial distal cutting (mm) Lateral distal cutting (mm) Medial posterior cutting (mm) Lateral posterior cutting (mm)
Group 1
Group 2
Group 3
Rectangular Gap (n = 64)
Tight Medial Gap (n = 41)
Tight Lateral Gap (n = 7)
2.2 ± 9.4 ± 9.2 ± 12.0 ± 9.6 ±
1.8 3.1 3.1 2.9 2.8
4.7 ± 10.0 ± 8.3 ± 13.3 ± 7.6 ±
1.9 † 3.8 3.0 3.2 2.7 †
3.4 ± 8.9 ± 9.1 ± 8.1 ± 8.8 ±
2.4 1.9 2.1 3.9 ‡ 1.8
P (ANOVA ) .000 * .637 .437 .000 * .005 *
*The 3 groups were compared using ANOVA, and statistical significance was accepted for P values less than .05. Intergroup differences were assessed using the post hoc Bonferroni test. †Group 2 required more ER of the femoral component (P = .000) and less lateral posterior femoral bone cutting (P = .004) than did group 1. ‡In group 3, medial posterior cutting amounts were smaller than in groups 1 and 2.
postoperatively. The mean (SD) knee ROMs improved from 115.8° (21.2°) preoperatively to 124.0° (30.5°) at 1 year and to 132.5° (7.4°) at 4 years postoperatively, and mean (SD) alignment of the lower extremity mechanical axis improved from a varus value of 10.6° (5.4°) preoperative to 1.2° of varus (1.9°) at 1 year and to 1.1° of varus (SD 1.8°) at 4 years postoperatively. Significant improvements in clinical outcomes were obtained for all parameters at 1 year postoperatively. With regard to outcomes at 4 years postoperatively, only WOMAC scores improved significantly as compared with results at 1 year postoperatively (P = .001; Table 1). Comparisons of Clinical Results and Intraoperative Parameters of Femoral Bone Cutting in Relation to Initial Mediolateral Gap Differences Patients were grouped according to initial mediolateral gap differences. At zero degrees of knee extension, there were 73 knees with a rectangular gap (group 1), 34 knees with a tight medial gap (group 2), and 5 knees
with a tight lateral gap (group 3). Radiographic evaluations revealed significant differences in lower extremity alignment according to initial gap differences. In particular, more varus alignment was observed in group 2 (mechanical axis, varus 2.22°) than in the other 2 groups at 1 year postoperatively (group 1, varus 3.24°; group 3, varus 0.04°; P = .001). However, no significant difference was observed between these 3 groups in terms of preoperative and postoperative mean ROM and HSS or WOMAC scores. Amounts of ER of the femoral component and of medial distal cutting and medial posterior cutting of femur were significantly different among the initial gap groups at zero degrees of knee extension. Intergroup differences were assessed using the post hoc Bonferroni test. Group 2 showed more ER of the femoral component than did group 1 (P = .032), more medial distal cutting than did groups 1 (P = .000) and 3 (P = .020), and more medial posterior cutting than did groups 1 (P = .006) or 3 (P = .001; Table 2). According to initial
Table 4. Comparison of Clinical and Radiologic Results According to Final Mediolateral Gap Differences
HSS score Preoperative Postoperative 1 y Postoperative 4 y WOMAC score Preoperative Postoperative 1 y Postoperative 4 y Extension lag (°) Preoperative Postoperative 1 y Postoperative 4 y Further flexion (°) Preoperative Postoperative 1 y Postoperative 4 y Mechanical axis (varus, °) Preoperative Postoperative 1 y Postoperative 4 y
Group A
Group B
Rectangular Gap in Both Angles (n = 105)
Nonrectangular Gap in Any Angle (n = 7)
P (t test) *
61.7 ± 8.8 93.4 ± 4.3 94.5 ± 8.1
59.7 ± 9.8 91.8 ± 5.1 90.8 ± 5.4
.591 .406 .319
77.2 ± 10.9 31.3 ± 5.9 28.5 ± 6.3
76.7 ± 16.8 32.5 ± 7.4 31.3 ± 2.3
.944 .639 .327
9.5 ± 7.4 0.8 ± 2.2 0.2 ± 1.1
16.4 ± 6.3 1.7 ± 2.6 2.0 ± 4.5
.018 .337 .414
130.5 ± 14.9 132.6 ± 6.0 132.5 ± 6.8
100 ± 32.7 119.2 ± 10.7 135.0 ± 5.0
.048 .027 .424
10.7 ± 5.2 1.2 ± 1.8 1.0 ± 1.8
9.8 ± 8.0 0.8 ± 3.5 2.4 ± 1.6
.795 .776 .463
*Statistical analysis was performed using the independent t test, and significance was accepted for P values less than .05.
588 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 mediolateral gap differences at 90° of knee flexion, ERs of the femoral component and medial and lateral posterior femoral cutting amounts were significantly different between the initial flexion gap groups. Group 2 required more ER of the femoral component (P = .000) and less lateral posterior femoral bone cutting (P = .004) than did group 1, and groups 1 and 2 required more medial posterior cutting than did group 3 (Table 3). Comparison of Clinical and Radiologic Results in Relation to Final Mediolateral Gap Differences In most cases, we obtained a balanced extension gap by precise and stepwise release in extension, and most flexion gaps were balanced by adjusting initial posterior femoral bony cutting and ligament release. Patients were allocated to the final rectangular or nonrectangular gap groups according to final mediolateral gap difference at zero degrees of knee extension and at 90° of knee flexion. There were 105 knees with rectangular gaps (group A), in which gaps were balanced at full extension and at 90° of knee flexion, and 7 knees with nonrectangular gaps (group B). However, clinical HSS and WOMAC scores and preoperative and postoperative mechanical axes of the lower extremities were no different in these 2 groups. Group A showed a significantly better flexion range preoperatively and at 1 year postoperatively (Table 4). With regard to intraoperative parameters, no differences were found between these 2 groups with respect to femoral cutting levels or ER of the femoral component. Complications Two patients experienced a superficial wound infection but responded to intravenous antibiotics. There were 3 cases of symptomatic deep-vein thrombosis, but all cured spontaneously. No knee showed radiolucency, osteolysis, subsidence, or loosening during follow-up.
Discussion Two surgical techniques can be used to establish ligament balance, namely, the classic flexion-extension gap balancing technique and the measured resection technique. According to the measured resection technique, the surgeon removes from the distal femur and proximal tibia the amount of bone being replaced by the prosthetic component. Therefore, distal femoral resection and posterior femoral condylar resection thicknesses are equal to those of the metallic femoral component, whereas proximal tibial resection approximates to the thickness of the tibial component. Soft tissue is balanced to obtain stable and symmetrical flexion-extension gaps, but this technique may have accompanying problems in imbalanced patients. To resolve an imbalanced gap, multiple additional bone cuts or soft tissue releases are needed after complete tibial and femoral bone cuts have been made. However,
redo multiple bone cuts or soft tissue releases could create more imbalance. Conversely, the classic flexion-extension gap balancing technique is based on the appropriate sizing of the femur with respect to the anteroposterior dimension. Posterior femoral condylar resection, therefore, varies as determined by this sizing of the femur. Proximal tibial resection is performed to remove approximately 2 mm of bone from the more deficient tibial plateau. These 2 resections determine the dimension of the flexion gap, and the extension gap is created to match these dimensions. If appropriate posterior soft tissue release has not been performed, this technique is more susceptible to erroneous distal femoral resection than the measured resection technique, and therefore, a number of surgeons who have traditionally used the classic flexion-extension gap balancing technique now favor a combination technique, referred to as the modified gap technique, which attempts to blend the measured resection and the classic flexionextension gap balancing techniques. In our study, the tibia-first technique was used at 0° in both the coronal and sagittal planes, to prepare extension gaps before posterior femoral condyle cutting (flexion gap preparation). Cutting the femur first can deprive the surgeon of the many choices that can be readily calculated using the computer, such as balancing of the knee using adjustable femoral cuts. For example, if a surgeon tries to prepare the flexion gap first, the shape and size of the flexion gap should be determined before axial alignment is adjusted. However, we consider this to be very difficult, and furthermore, this difficulty is compounded when the knee has an advanced deformity. Accordingly, we advocate an alternative method, namely, that the rectangular extension gap be prepared first and that this be followed by an equal flexion gap. Navigation systems now provide femoral planning based on initial flexion and extension gap measurements. Based on gap differences, distal femur cutting and posterior condylar cutting depth can be planned and femoral component size and rotation adjustments can be simulated to achieve flexion and extension gap balance. Furthermore, final extension gaps can be adjusted during navigation-assisted TKA by modifying the distal femur cutting depth and flexion gap configuration. During this stage, we did not adjust the coronal cutting angle to obtain cutting perpendicular to the femoral mechanical axis, and thus, soft tissue release is most important task in terms of providing a neutral coronal limb axis. As can be done during the modified gap technique, the ER of the femoral component can be adjusted to obtain a balanced flexion gap, mediolaterally. However, in cases with excessive ER of the femoral component, the specific portion of soft tissue responsible for the tight flexion gap must be released to avoid patellofemoral problem.
Rectangular Flexion and Extension Gaps During Primary TKA Seon et al
Until recently, intraoperative gap balancing measurement has depended on subjective assessments made by operating surgeons. However, in the present study, we used a calibrated torque wrench and tensor system as a means of obtaining reproducible measures of gap differences, and we believe that this method could be used to help surgeons determine optimal balance in individual knees earlier during operations. Unitt et al [11] measured flexion-extension gaps in 218 TKAs using the measured resection technique before and after soft tissue release and defined a nonbalanced gap as one with more than 3 mm of laxity. Based on their results, balanced flexion and extension gaps during TKA were achieved using the measured resection technique in 175 knees (80.3%). In the present study, we obtained a final rectangular gap in 105 knees (94%) using the navigation-assisted gap balancing technique. When we compared the results of the present study with those of the study of Unitt et al, we found that TKA using the navigation-assisted gap balancing technique produced better balanced flexion and extension gaps than TKA using the measured resection technique. Relatively few studies have addressed the influence of soft-tissue balancing on TKA outcomes [11-13]. Attfield et al [12] reported that knees balanced in both full extension and in flexion showed improved proprioception postoperatively, and in the present study, significant improvements in clinical outcomes were observed at 1 year postoperatively after TKA. However, only WOMAC scores showed a significant improvement at 4 years postoperatively versus 1 year postoperatively. Lower extremity alignments and good clinical outcomes were also maintained at 4 years after TKA. However, no association was found between final gap differences and preoperative to postoperative differences in terms of mechanical axis alignments or clinical outcomes, such as HSS and WOMAC scores. However, although we tried to achieve a rectangular gap, we failed to do so in 7 patients, which were particularly true for flexion gaps, although no patient required a constrained prosthesis. The minimum gap difference in the nonrectangular gap group was 5 mm, and even though a rectangular final gaps were not achieved when measurements were conducted using a tensioner, knees were within the acceptable range of instability clinically by manual testing, and thus, results in these cases were accepted. This may explain the lack of a difference between the outcomes of balanced and imbalanced knees. However, patients with greater preoperative flexion ranges were found to be more likely to have a rectangular final mediolateral gap than a nonrectangular gap. Furthermore, we believe that it was easier to achieve rectangular gaps in knees with less deformity, although by using the navigationassisted modified gap balancing technique, we were able to obtain a rectangular gap at similar rates in
589
coronally deformed and less deformed knees. With regard to ROM, those with well-flexed knees tended to have greater flexion gaps during surgery, which resulted in greater flex postoperatively if a wellbalanced gap was obtained during index surgery. Navigation systems are being increasingly used for TKA and now provide excellent mechanical axis restoration and precise component positioning. In addition, these improvements increase the accuracy of the balancing procedure by providing more objective and quantitative measures of flexion and extension gaps [6]. In the present study, we defined a rectangular gap as one with a flexion-extension difference of 3 mm or less, a tight flexion gap as one with an extension gap of 3 mm more than the corresponding flexion gap, and a tight extension gap as one with a flexion gap of 3 mm more than the corresponding extension gap. To date, no study has been undertaken to determine the relation between the amount of femoral bone cut and initial and final gap differences. In the present study, we analyzed relations between initial gap differences and the amounts of femoral bone cut and the ERs of femoral components. In the tight medial gap group at 0° of knee extension, more medial side distal and posterior femoral bone cutting were needed to balance final gaps. This suggests that degrees of femoral bone cutting and ER can be predicted from initial gap differences after proximal tibial bone cutting and preliminary soft tissue release. In the present study, medial posterior femoral bone cutting amounts were found to be related to initial gap difference at 90° of flexion and at 0° of knee extension. Furthermore, the ERs of femoral components were found to be more related with initial gap differences at 90° of flexion than in extension. In addition, the tight medial gap group at 90° of flexion needed a mean 2.5° of ER more than the balanced group. In the rectangular gap group, amounts of bone cutting and the ER of femoral implants during the gap balancing technique were similar to those of the measured resection technique. In terms of the strength of this study, we provided the surgeon with information regarding the amount of bone cutting required based on initial gaps, and we attempted to create a rectangular space using same tensions during gap balancing. However, the limitations of this study are that it does not contain a control group or a nonnavigated group and that the final nonrectangular gap group was too small to provide comparative data. Further study is needed to compare the clinical and radiologic outcomes of patients with balanced and imbalanced gaps. In conclusion, amount of intraoperative femoral bone cutting and the ERs of femoral components were found to depend on initial gap differences. However, no differences in clinical or radiologic outcomes were observed between the final rectangular and nonrectangular groups at 1 or 4 years after TKA. Furthermore,
590 The Journal of Arthroplasty Vol. 26 No. 4 June 2011 navigation was found to provide information about the amount of bone cutting required based on initial gap differences. In addition, we found that navigation provides an effective means of achieving rectangular flexion-extension gaps during TKA performed using the navigation-assisted modified gap balancing technique.
References 1. Freeman MA, Todd RC, Bamert P, et al. ICLH Arthroplasty of the knee: 1968-1977. J bone Joint Surg Br 1978;660-B: 339. 2. Insall JN, Binazzi R, Soudry M, et al. Total knee arthroplasty. Clin Orthop 1985;192:13. 3. Fehring TK, Odum S, Griffin WL, et al. Early failure in total knee arthroplasty. Clin Orthop Relat Res 2001;392:315. 4. Sharkey PF, Hozack WJ, Rothman RH, et al. Insall award paper. Why are total knee arthroplasties failing today? Clin Orthop Relat Res 2002;404:7. 5. Dorr LD, Boiardo RA. Technical considerations in total knee arthroplasty. Clin Orthop 1986;205:5. 6. Clemens U, Miehlke RK. Advanced navigation planning including soft tissue management. Orthopedics 2005;28 (10 Suppl):s1259.
7. Matsumoto T, Muratsu H, Tsumura N, et al. Soft tissue balance measurement in posterior-stabilized total knee arthroplasty with a navigation system. J Arthroplasty 2009;24:358. 8. Han SB, Nha KW, Yoon JR, et al. The reliability of navigation-guided gap technique in total knee arthroplasty. Orthopedics 2008;31(10 Supp1) pii: orthosupersite. com/view.asp?rID=35542. 9. Higuchi H, Hatayama K, Shimizu M, et al. Relationship between joint gap difference and range of motion in total knee arthroplasty: a prospective randomized study between different platforms. Int Orthop 2009;33:997. 10. Romero J, Stahelin T, Binkert C, et al. The clinical consequences of flexion gap asymmetry in total knee arthroplasty. J Arthroplasty 2007;22:235. 11. Unitt L, Sambatakakis A, Johnstone D, et al. Short-term outcome in total knee replacement after soft-tissue release and balancing. J Bone Joint Surg Br 2008;90-B:159. 12. Attfield SF, Wilton TJ, Pratt DJ, et al. Soft tissue balance and recovery of propioception after total knee replacement. J Bone Joint Surg Br 1996;78-B:540. 13. Mihalko WM, Whiteside LA, Krackow KA. Comparison of ligament balancing techniques during total knee arthroplasty. J Bone Joint Surg Am 2003;85-A(Suppl 4):132.