Abstract
Valgus-producing high tibial osteotomy (HTO) is an established and effective treatment for varus malalignment and medial gonarthrosis, with good to very good results in long-term studies, especially with a slight overcorrection of the mechanical axis. § The survival rate of the procedure can be improved if special attention is paid to proper patient selection with regard to age, body mass index, and knee joint range of motion. 2,20,25 Long-term joint preservation is based on the effect of valgus HTO in the frontal plane, transferring the mechanical load from the overloaded and worn medial knee compartment to the healthier lateral knee compartment.
Apart from the mentioned and well-investigated effect of valgus HTO in the frontal plane, more and more attention has been paid to the effect in the sagittal plane on posterior tibial slope (PTS) and patellar height (PH). In a recent study by some of the current authors, PTS increased by 2° to 5° after open-wedge HTO and decreased by 2° to 5° after closed-wedge HTO. 5,10 These values are similar to those found by other authors. 5,15,17,18,19 Thus, unintentional increases in PTS are a well-known consequence of open-wedge HTO.
Previous studies have also shown that PH decreases with open-wedge HTO 3 and increases with closed-wedge HTO, 3 without a strong correlation to the degree of valgus correction in the coronal plane. 7,9,17 Most of these studies are retrospective, and HTO was performed as a coronal plane procedure without explicitly addressing the sagittal changes. Recent studies stated that the typical PTS increase in valgus open-wedge HTO cannot be controlled if the plate is anteromedial or centromedial. 17,24 There is only one study that tried to control PTS and PH by osteotomizing above or below the patellar tendon attachment in valgus open-wedge HTO. 19
Therefore, the current hypotheses were as follows:
Modifications of surgical technique details can prevent changes in tibial slope.
By specifically addressing the tuberosity fragment in relation to the amount of correction, PH can be left unchanged.
Materials and Methods
All patients agreed to participate in the study. The patient consent procedure strictly involved disclosure, understanding, voluntariness, and competence. The Institutional Review Board did not authorize use of a CT scan for postoperative imaging.
Patient Data
The study consisted of 25 knees in 25 consecutive patients who had an open-wedge HTO performed by one of the authors (S.H.) between August 2008 and May 2009 for varus malalignment and medial compartment disease (ie, localized cartilage damage in the medial compartment or subtotal resection of the medial meniscus in previous surgery, each with medial pain but good motion). There were 3 women (12%) and 22 men (88%), with a mean age of 40.2 ± 8.9 years at surgery. Because the indications were not for osteoarthritis and the patients were young, the average correction was less than the recommended 3° to 4° of mechanical valgus. All patients agreed to participate in this study. High tibial osteotomy was a single procedure in 17 knees (65%), combined with osteochondral transplantation in 2 knees (8%), microfracturing in 3 knees (12%), and implantation of a collagen meniscus implant in 3 knees (12%). Combined 1-stage double-bundle anterior cruciate ligament replacement was performed in 2 patients with varus malalignment and anterior cruciate ligament insufficiency. One of these patients also required microfracturing of the medial femoral condyle.
Diagnostics
The preoperative diagnostic studies consisted of conventional anteroposterior (AP) and lateral radiographs, a full-length standing AP radiograph of the affected leg, and an MRI of the affected knee. Leg alignment was determined with the long standing films (ie, mechanical varus of the lower extremity, as given by the hip-knee-ankle axis 13 ).
Surgical Procedure
The operation was performed with the patient in the supine position on a straight operating table. The procedure was started with arthroscopy to define the cartilage status in the patellofemoral and medial and lateral femorotibial compartments. Then, after the proximal tibia was exposed through a 10-cm anteromedial incision, 2 reference K-wires were drilled in the sagittal plane, 1 above the osteotomy directly proximal to the insertion of the patellar ligament and a second 8 cm below the osteotomy. These K-wires were parallel in the sagittal and axial planes. The sartorius fascia was incised proximal to the gracilis tendon. The incision went from the tibia to the ventral portion of the superficial medial collateral ligament. So, the gracilis and semitendinosus tendons were exposed and mobilized, and the gracilis attachment was detached from bone. The ventral portion of the superficial medial collateral ligament was incised. By retaining the medial collateral ligament dorsally and the gracilis and semitendinosus tendons distally, the level of the osteotomy was exposed, starting 4 to 5 cm below the joint line.
The osteotomy was obliquely aimed upward in the direction of the fibular head, about 2 cm below the lateral joint line, and was gradually opened to preserve a lateral bony bridge. The osteotomy was performed in a biplanar fashion, leaving the tibial tuberosity on the distal fragment if there were no patellofemoral complaints preoperatively and the correction did not exceed 5° (13 patients, 52%, 37.8 ± 7.8 years). Preoperative complaints included, from the medical history, pain behind the patella during sports or activities of daily living (eg, walking, climbing stairs) or, from the physical examination, pain behind the patella on pressure or mediolateral translation under pressure. Otherwise, the frontal plane osteotomy was aimed distally, leaving the tuberosity on the proximal fragment to avoid altering patellofemoral kinematics (12 patients, 48%, 42.8 ± 9.6 years). To avoid the increase in PTS routinely seen with open-wedge HTO, the recommendations of Hernigou et al 13,14 and Noyes et al 22 were followed: First, a complete posterior osteotomy was performed, and the spreader for opening the osteotomy was placed as far posterior as possible (ie, as close to the posterior medial corner of the tibia as possible). Additionally, the leg was positioned in a sagging position, and the opening gap at the tibial tubercle was kept one-half the dimension of the posteromedial gap. During the procedure, the 2 reference sagittal K-wires were kept parallel to ensure maintenance of the original tibial slope. Finally, the open-wedge osteotomy was fixed with a self-locking plate (Tomofix, Synthes, Umkirch, Germany), and the tuberosity was fixed with 1 or 2 screws if the osteotomy was aimed distally. The position of the Tomofix plate was exactly adjacent and thus slightly anterior to the posteriorly placed spreader. Figure 1 shows and describes the preoperative and postoperative radiographs, as well as the intraoperative photographs.
Two sets of anteroposterior and lateral radiographs of right knees and intraoperative photographs. A, the 2 films on the left are the preoperative films of a slightly varus knee with medial osteoarthritis in which a double-bundle anterior cruciate ligament reconstruction has been performed. The films on the right show the completed biplanar osteotomy with a proximal tubercle osteotomy. The high tibial osteotomy is fixed with a self-locking plate. B, the 2 films on the left are the preoperative films of a moderately varus knee with medial osteoarthritis. The films on the right show a larger biplane open-wedge high tibial osteotomy with a distal tubercle osteotomy. The tubercle osteotomy has been fixed with 2 screws, and the high tibial osteotomy, with a self-locking plate. C, the intraoperative photographs show the following: top, the marking of the osteotomies done by diathermia, with the detachment of the gracilis tendon at the junction of the tibial head osteotomy and the tubercle osteotomy (the 2 reference K-wires are parallel); middle, osteotomy performed in a biplanar fashion, leaving the tibial tuberosity on the proximal fragment (the spreader for opening the osteotomy is placed as far posterior as possible; the clamp marks the posterior border of the tibia); bottom, the open-wedge osteotomy, as fixed with a self-locking plate, and the tuberosity osteotomy aiming distally, as fixed with 2 screws (the 2 reference K-wires are still parallel).
Imaging
Anteroposterior and lateral radiographs were obtained the first day after the operation. To obtain true AP radiographs of the knee, care was taken to make sure that the patella was pointing anteriorly. Special attention was paid to ensure that the lateral radiographs were taken with the knee in 30° of flexion and the limb in neutral rotation. The beam was centered over the joint line, with the goal of superimposing the femoral condyles but accepting rotation or tilting up to 5 mm. If the radiographs were unsatisfactory, they were retaken.
Frontal Plane Correction Measurement
Postoperatively, the magnitude of correction in the frontal plane was determined by measuring the angle between the proximal and distal border of the osteotomy on the postoperative AP radiograph using digital X-ray software (PACS, Philips EasyVision, Best, Netherlands). All measurements were done digitally on the computer by one of the authors (K.B.). The radiographs were magnified to increase accuracy of line placement.
Slope Measurement
On the lateral radiograph, PTS for the medial and lateral tibial compartment was measured in relation to the posterior tibial cortex, a line whose use is supported by the research of Genin et al 12 and the method described by Brazier et al. 4 Figure 2B describes the reference lines and technique.
PH Measurement
Patellar height was assessed on the preoperative and postoperative lateral radiographs using the Caton-Deschamps Index. 6 See Figure 2C for the reference lines and the formula for calculation. The index ratio for normal PH is between 0.6 and 1.3. Patella infera is diagnosed if the index is equal to or below 0.6. Patella alta is diagnosed if the index is equal to or more than 1.3.
A, normal cross-table lateral radiography of a right knee flexed to 30°. Note that the femoral condyles are perfectly superimposed. B, posterior tibial slope measurement using the method of Brazier et al. 4 Posterior tibial slope was determined by digitally measuring the angle between the medial tibial plateau (MTP) and perpendicular to the posterior tibial cortex (PTC; 15 cm of proximal tibia). The same is done for the lateral tibial plateau (LTP). C, patellar height was assessed with the Caton-Deschamps Index 6 —that is, by dividing AT (distance from the inferior border of the patellar articular surface to the proximal anterior tibial rim) by PA (height of patellar articular surface).
Statistical Analysis
The statistical software used for all analyses was SPSS 15.0 (SPSS Inc, Chicago, Illinois). We used a paired t test for comparison of preoperative and postoperative PTS and PH. A P value of less than .05 was regarded as statistical significance.
Results
Table 1 summarizes the results.
Values and Statistical Analysis: Before and After Open-Wedge High Tibial Osteotomy
With frontal-plane high tibial osteotomy correction.
Patella alta: n = 0. Patella infera: n = 2.
Patella alta: n = 0. Patella infera: n = 0.
Patient Data
The full standing AP radiographs revealed a mean varus deformity of 5.2° ± 2.8°. Mean correction angle in the coronal plane was 7.4° ± 2.6°. Because the indications were not for osteoarthritis and the patients were so young, the average correction was less than the recommended 3° to 4° of mechanical valgus.
Posterior Tibial Slope
The average preoperative medial and lateral PTS was 4.2° ± 2.9° and 5.4° ± 3.0°, respectively. The postoperative medial and lateral PTS was 4.2° ± 3.9° and 5.1° ± 2.7°, respectively. The postoperative medial PTS was not significantly different from the preoperative (P = .49); the average increase accounted for 0.1° ± 4.5°. Similarly, the postoperative lateral PTS showed no significant difference from the preoperative (P = .48); the average increase accounted for −0.3° ± 3.7°.
Patellar Height
According to the Caton-Deschamps Index, the mean PH ratios increased from 0.92 ± 0.15 preoperatively to 0.97 ± 0.19 postoperatively. This increase was not statistically different. In the group of proximal tuberosity osteotomy, these figures were 0.95 ± 0.17 and 0.98 ± 0.24, respectively (P = .53). In the group of distal tuberosity osteotomy, the corresponding figures were 0.89 ± 0.13 and 0.95 ± 0.14 (P = .266). No patient showed patella infera preoperatively or postoperatively. There was no correlation between PH changes and HTO correction degrees.
Discussion
The aim of this prospective study was to evaluate changes in tibial slope and PH that occur during valgus-producing open-wedge HTO when techniques to prevent these changes are employed: spreader placed far posterior, as close to the posterior medial corner of the tibia as possible; leg in a sagging position; opening gap at the tibial tubercle one-half the dimension of the posteromedial gap; parallel position of the sagittal K-wires during the procedure. This study proved the hypothesis that PTS does not change if the slope is carefully controlled with a combination of old and new techniques. The study also proved the hypothesis that significant changes in PH can be avoided by employing a biplanar osteotomy that leaves the patellar tendon on either the distal or proximal fragment, depending on the patellofemoral findings—namely, tibial tuberosity on the distal fragment if there is pain behind the patella during activities of daily living or sports, pain behind the patella on pressure or mediolateral translation under pressure, correction exceeding 5°.
The authors of prior studies have shown that there is an average increase in tibial slope of 2° to 5° if open-wedge HTO is performed 9,10 and no attempt is made to control PH. Other authors have shown similar slope changes. 5,7,17,19,28 Using the gap principle of Noyes et al, 22 adhering to the surgical methods of Hernigou et al, 13,14 and employing other helpful techniques, we sought to eliminate changes in PTS. In the current study, the slope was monitored with parallel pins placed in the sagittal plane; the soft tissues were stripped off the posterior tibia 25 ; the osteotomy was completed posterolaterally 27 ; the anterior osteotomy gap was half the posteromedial gap 22 ; and the fixation was applied as far posteriorly as possible.
Additionally, we placed the spreader for opening the osteotomy as far posterior as possible. According to the technique by Sariali and Catonne, 24 slow posterior spreading was performed with gradually increasing-sized wedges to preserve the lateral hinge. Furthermore, the leg was positioned in a sagging position while plate fixation was performed. As a result of careful attention to the prevention of slope change, the current study showed no significant difference between the preoperative and postoperative medial and lateral PTS. We interpret that as another sign that the osteotomy was perpendicular to the posterior cortex in the axial plane so that the lateral hinge was perfectly lateral. Thus, the unwanted negative effects that occur with PTS changes can be minimized, such as limited range of motion, worsening of knee instability, progression of osteoarthritis, or complications of future surgeries (eg, ligament reconstructions, knee replacements). 10,13,16
Concerning the second sagittal plane parameter, PH decreases with open-wedge HTO 7,9 because the patellar tendon remains on the distal fragment and moves distally in relationship to the joint line when the osteotomy is opened. The patella is thus lowered in relationship to the joint line. In the current study, a biplanar osteotomy was performed with the tibial tuberosity left on the distal fragment (in small coronal corrections) and the proximal fragment (in patients with patellofemoral complaints preoperatively and/or with a correction exceeding 5°); as such, PH as defined by the Caton-Deschamps Index increased only 0.03 and 0.06, respectively. The distal osteotomy technique of the current study uses the same principle as that of Matar et al, 19 who placed the osteotomy at or distal to the tibial tubercle, leaving the patellar tendon attached to the proximal fragment. Again, potential troublesome changes for the patellofemoral joint can be minimized—especially those associated with patella infera, such as altered kinematics, pain syndromes, and complications for future surgeries (eg, patellar mechanism eversion difficulties during total knee replacement). 3,9
One of the defining characteristics of the current study is that it carefully tried to prevent the slope changes routinely seen with the performance of open-wedge HTO. We followed and modified the above-mentioned rules by Hernigou et al 13,14 and Noyes et al 22 and thus did not see PTS increase by 2° to 8° after open-wedge HTO. 10,17,21 Latest publications with a focus on that topic confirm our above-mentioned approach, saying that the typical slope increase in valgus open-wedge HTO can be controlled with only a locked implant placed at the posteromedial cortex. 24 Positioning a nonlocked implant at the posteromedial cortex might also result in slight PTS increase. 16 Finally, we must discuss why the initial PTS values in this study are lower than what the literature indicates. 22 We think that the posterior tibial cortex method that we use gives the lowest PTS results anyway. Furthermore, the mean age of all the patients investigated in the literature is much higher than that of our study group, and those patients showed more severe stages of osteoarthritis than ours did. 22
In a study by Wright et al, 28 64% of the patients satisfied the radiographic criteria for patella infera after open-wedge HTO, according to the Blackburn Peel Index. The Insall Salvati Index decreased insignificantly in 29% of patients, and the Blackburn Peel Index decreased significantly from 0.75 to 0.53. In a recent study by Chae et al, 7 26% of cases satisfied the radiographic criterion for patellar infera after open-wedge HTO, according to Blackburn Peel Index. The mean index significantly declined from 0.71 to 0.61. Because our surgical procedure strictly followed the rules by Hernigou et al 13,14 and Noyes et al, 22 we saw no significant changes in PH according to the Caton-Deschamps Index. We also saw this in the group of distal osteotomies, which were performed in the case of additional patellofemoral complaints.
Our current results concur with our prior work 10 in that PTS changes are not dependent on the amount of frontal plane correction. As far as PH is concerned, our results concur with our recently published data 9 indicating that PH changes are not dependent on the amount of frontal plane correction.
The current study focused mainly on geometric values. It lacks the correlation between clinical outcome and changes in PTS and PH. This subject should be studied in a future well-planned prospective study comparing clinical outcome and geometric changes. The measurement techniques for PTS and PH are affected by the rotation of the tibia in the lateral view when measured with a conventional radiographic technique. 16 We addressed this problem by using strictly lateral radiographic views with good superimposition of the posterior condyle contours. An even more precise method to measure medial and lateral PTS would be a CT scan or MRI. The local ethics committee did not authorize use of a CT scan, whereas MRI led to intolerable artifacts owing to the osteosynthesis material in the postoperative setting. We used the Caton-Deschamps index to measure PH, knowing that this technique might be influenced by changes in tibial slope. However, because we reached our goal not to change PTS, we could confidently use that well-known and accepted index without any systematic mistake.
We previously noted internal rotation of the distal tibia while doing open-wedge HTO. For a pilot study, we measured the axial rotation of the pins used to control PTS and noted 5° of internal rotation of the distal tibia in the 23 patients in whom the gracilis and semitendinosus tendon was not removed. Axial rotation during open-wedge HTO should be the subject of a future study.
Conclusion
This study shows that PTS and PH in valgus open-wedge HTO can both be controlled by use of specific intraoperative techniques. These methods enable the surgeon to perform open-wedge HTO for medial compartment disease and varus deformity with precision in not only the coronal plane but the sagittal plane as well. With control of PTS, the extension deficit commonly seen with open-wedge HTO can be eliminated, and with the appropriate osteotomy of the tibial tubercle during open-wedge HTO, the negative effects on the patellofemoral joint can be minimized.
