Three-column Subdivision for Isolated Posterolateral Tibial Plateau Fractures and Perspective Surgical Approachs

Management of the tibial plateau fracture involving posterolateral compartment is technically challenging. This study aimed to introduce a computed tomography-based classication of the posterolateral compartment of tibial plateaus based on the bula and to propose the individualized surgical approaches for each zone; and 2) to determine the surgical approach based on the classication, would achieve a safe and effective reduction and can improve postoperative clinical outcomes. Eighteen cases of tibia plateau fracture involved the isolated posterolateral compartment in a single institution were retrospectively analyzed. The posterolateral compartment of tibial plateau was segmented into three zones based on the bular position and individualized surgical approach was proposed for each zone. In anterior Zone I, surgical treatment was performed using an extended anterolateral approach and patient was placed in the supine position; In middle Zone II, using the transbular approach in the supine position; In posterior Zone III, using the posteromedial approach in prone. Articular reduction (intra-articular step off in CT images) and mechanical medial proximal tibial angle (mMPTA) in simple radiographs taken in the immediate postoperative period and clinical outcomes of the Lysholm knee scoring scale and range of motion were evaluated at postoperative 1-year follow-up, of tibial plateau fractures. Several approaches and reduction techniques


Abstract
Background Management of the tibial plateau fracture involving posterolateral compartment is technically challenging. This study aimed to introduce a computed tomography-based classi cation of the posterolateral compartment of tibial plateaus based on the bula and to propose the individualized surgical approaches for each zone; and 2) to determine the surgical approach based on the classi cation, would achieve a safe and effective reduction and can improve postoperative clinical outcomes.

Methods
Eighteen cases of tibia plateau fracture involved the isolated posterolateral compartment in a single institution were retrospectively analyzed. The posterolateral compartment of tibial plateau was segmented into three zones based on the bular position and individualized surgical approach was proposed for each zone. In anterior Zone I, surgical treatment was performed using an extended anterolateral approach and patient was placed in the supine position; In middle Zone II, using the trans bular approach in the supine position; In posterior Zone III, using the posteromedial approach in prone. Articular reduction (intra-articular step off in CT images) and mechanical medial proximal tibial angle (mMPTA) in simple radiographs taken in the immediate postoperative period and clinical outcomes of the Lysholm knee scoring scale and range of motion were evaluated at postoperative 1-year follow-up,

Results
In all cases, anatomical articular reduction (intra-articular step off < 2 mm) was achieved, maintained for the follow-up period. The average mMPTA was increased from 87.6º before surgery to 88.2º in immediate postoperative period (p = 0.060), and maintained for the follow-up period (mean 89.9º at one-year followup). At the 1-year follow-up, the knee range of motion averaged 140 degrees and the Lysholm knee function scored 95.0 points.

Conclusion
An individualized surgical approach and xation according to three-zone subdivision for isolated posterolateral tibial plateau fractures provided an effective and safe method to treat posterolateral tibial plateau fractures.
Level of evidence: Therapeutic study, IV Background Tibial plateau fracture involve the articular surface, should be anatomically reduced to minimize complications such as arthritis and angular deformation. Isolated posterolateral tibial plateau fractures account for about 7-10% of all tibial plateau fractures. Several approaches and reduction techniques have been introduced to prevent the neurovascular injury in close proximity to the posterior nerve and vessels during surgical treatment and to overcome the technical di culty of the access and reduction to fracture site by proximal bula [1].
However, respective approaches for posterolateral tibial plateau fracture surgery have their pros and cons. The anterior and anterolateral approach has the advantage of easy access to the joint, while the displaced articular surface can be xed with no possibility of direct fracture manipulation. In that sense, the anterolateral approach combined with the bular osteotomy is advantageous to identify a large area of the articular surface including the posterolateral area with direct vision. However, there would be a risk of nonunion of the osteotomed bula or the peroneal nerve injury. The posterior or posterolateral approach has the technical advantage to enable to position the posterior buttress plate and to reduce the posteriorly incarcerated and depressed bony fragments with direct vision to the posterior articular surface. However, the posterolateral approach itself may injure the posterolateral ligament complex of the knee and can result in knee instability and has a possibility of iatrogenic injury to the popliteal artery and nerve [2].
Therefore, preoperative planning based on individual fracture patterns and con gurations is important to achieve the accurate reduction and to minimize postoperative complications. Several authors have segmented the tibial plateau and proposed respective surgical approachs and xation methods accordingly. Luo et al. proposed "Three-column concept" for the analysis of the fracture pattern of tibial plateau. They suggested individual surgical approach for each compartment, the posteromedial approach for medial and posterior compartment fractures, the anterolateral approach for lateral compartment fracture [3].
However, the previous "Three-column concept" did not resolve which surgical approach and reduction technique would be appropriate for the tibial plateau fracture involving posterolateral compartment. The structures at the posterolateral corner region such as proximal bula and lateral collateral ligament obscure to visualize the posterolateral articular surfaces and to allow access to the space. Therefore, Hoekstra et al. proposed a revised three-part classi cation that additionally classi ed the posterolateral corner region based on the bula region [4]. Cho et al. introduced a modi ed posterolateral approach that allows access to the space around the lateral collateral ligament which between the lateral collateral ligament and the lateral plateau margin, exposing the lateral and posterolateral articular surfaces, and reducing them [5].
In this study, the posterolateral compartment of tibial plateau was divided into three zones based on the bular head position and individualized surgical approaches were proposed for each zone. It was hypothesized that the individual understanding of fracture pattern and con guration at the posterolateral tibial plateau fractures, would be useful to determine the preoperative planning as surgical approach, and our strategy would provide clinical applicability to achieve a safe and effective reduction and to improve postoperative clinical outcomes.

Patient population
A retrospective analysis was performed on 42 cases of tibia plateau fracture involved the posterolateral compartment in a single institution for a total of 7 years from 2011 to 2017. All patients with tibia plateau fractures were studied by CT scans and classi ed by AO/OA classi cation and three column classi cation by Luo et al. [3]. Out of this 42 cases, isolated posterolateral tibial plateau fractures account for 18 cases, about 8.3% (18/215 cases) of all tibial plateau fractures. All 18 patients were operated and followed for more than one year (Fig. 1.). This study was approved by the institutional ethics committee at our hospital and got the IRB approval.
Three-part classi cation of tibial plateau fracture involving isolated posterolateral compartment.
The posterolateral compartment of the tibial plateau was divided into three zones centered on the bula. Individual surgical approach and operative techniques based on zone classi cation All surgeries were performed by one experienced orthopedic surgeon and surgical approach was differently chosen according to where is the most displaced area in posterolateral compartment. When the con guration of fracture showed most displaced in Zone I ( Fig. 2-b), treated using an extended anterolateral approach. The patient was placed in the supine position, the tourniquet was applied, the location of the bula and the joint surface were checked, and an inverted L-shaped curved incision was made. Then, the iliotibial ligament was detached from the Gerdy's tubercle ( Fig. 3-a) to expose the posterolateral part of the tibia ( Fig. 3-b). The articular surface can be visualized with enough space for direct fracture manipulation ( Fig. 3-c).
In Zone II ( Fig. 2-c), treated using the trans-bular approach with bular head osteotomy [6]. The patient was placed on the supine position, the tourniquet was applied, and the thigh holder was used for exion of the knee. The border between the thigh biceps tendon and bular head was checked, and a longitudinal incision was made in front of the thigh biceps tendon. (Fig. 4-a). After con rming the femoral biceps tendon and the iliotibial ligament, the femoral fascia was dissected. After identi cation and protection the common peroneal nerve, the bular head osteotomy was done ( Fig. 4-b). The articular surface was xed with possibility of direct visualization (Fig. 4-c and d). The bular head was re-xed using the tension band wire after xation.
In Zone III ( Fig. 2-c), treated using the posteromedial approach [7]. The patient was placed on prone position and the tourniquet was applied. The knee was exed about 20 degrees, and an inverted "L" shaped incision, starting from the parallel popliteal fold, was bent from the medial to the distal. Blunt dissection performed to expose the medial head of the gastrocnemius muscle, which was retracted with important neurovascular structures. After exposing the fractured site by blunt dissection between the soleus and the retracted medial head of the gastrocnemius ( Fig. 5-a), the displaced fragment was reduced and xed using posterior buttress plate under the direct vision ( Fig. 5-b and c).

Radiologic and clinical outcome evaluations
Postoperatively, anteroposterior X-rays of the knee were taken in the immediate postoperative period, 6 weeks, 12 weeks, and every 6 to 8 weeks until bony union occurred and then one year after the index operation. The radiographic evaluations entailed full-length, standing hip-to-ankle, knee AP, and lateral and patellar Merchant radiography. Before acquiring each radiograph, a radiology technician ensured that neutral alignment was achieved using the method described by Paley [8] to achieve a true AP image. This involved maintaining a forward knee position with the patella centered on the femoral condyles. To control the rotational position during the two radiographic examinations, an identical neutral foot rotation angle was obtained using a reference template on the platform incorporated for the plain radiographic system. In addition, during both radiographic examinations, the patients were requested to maintain knee extension to get an identical knee angle. The positions of the weight-bearing line (WBL), Mechanical femorotibial angle (MFTA), and mechanical medial proximal tibial angle (mMPTA) were measured on fulllength, standing hip-to-ankle radiography in the postoperative 6-month and 1-year after surgery [9]. WBL was de ned as the distance (%) of the total width of the tibia from the innermost point of the tibial plateau to the point where the mechanical axis passes. The MFTA was de ned as the angle between the mechanical axes of the femur and the tibia, and the mMPTA as the medial angle between a line drawn parallel to the proximal tibial condyles and the mechanical axis of the tibia [10]. An independent investigator drew these angles on postoperative full length radiographs and compared the values between the sides.
Within a week after operation, CT scan images were acquired to evaluate whether the accurate restoration of depressed articular surface was achieved and the appropriate xation was attained with no reduction loss in the immediate postoperative period and at postoperative one-year follow-up. Anatomical reduction of the articular surface was de ned as intra-articular step off of less than 2 mm. Secondary loss of reduction was de ned as an articular depression of 2 mm or more when compared with the rst postoperative CT scan at postoperative one-year follow-up [3] (Fig. 6).
To con rm the functional outcomes after the surgery, the Lysholm knee scoring scale was evaluated from 3 months postoperatively, when the weight bearing and gait were allowed, followed by 6-month and 1- year follow-up. The range of motion of the knee joint was also evaluated at the outpatient follow-up.

Statistical analysis
All data analysis was done using SPSS 20.0 (SPSS Inc., Chicago IL). Descriptive statistics were used to determine ranges, means, and standard deviations. Binary outcomes for the two groups were analyzed using Fisher's exact test and continuous outcomes were analyzed with the unpaired t-test. Pre and postoperative changes were evaluated by the paired T-test. Correlations were analyzed by using the Pearson correlation coe cient. P < 0.05 was considered statistically signi cant.

Results
All 18 patients were available at the one-year follow-up (mean, 19.8 months). Of the 18 patients, 12 were male and 6 were female. The mean age was 52.1 years old. The details of patients' demographic data were presented in Table 1. Most of the surgeries were performed using individualized approach according to zone classi cation, except for one case with a window through fracture site (Table 2). Nine (50%) of the 18 patients had fractures con ned to zone I fractures. The reduction and xation were performed using an extended anterolateral approach. Three patients (17%) had fractures con ned to the zone II fracture and were reduced and xed using the trans-bular approach with bular head osteotomy. Two cases (11%) were found to have fractures extending from the zone II to the zone I, and one case was using the trans-bular approach and the extended anterolateral approach. One case was using an extended anterolateral approach and window through fracture site. One case (6%) was con ned to the third zone fracture and the reduction and xation were performed using the posteromedial approach. One case (6%) was found to have fractures extending from the zone III to the zone I and zone II. Reduction and xation were performed using the posteromedial approach and the extended anterolateral approach (Table 3).    Table 3).
The knee range of motion averaged 140 degrees (ranged from 0 degrees of extension to 150 degrees exion) at the 1-year follow-up and the average Lysholm knee function score was 95.0 points at the oneyear follow-up.

Discussion
With the strategy of individualized surgical approach and manipulation of displaced fracture, based on the division into three zones based on the bular head position, the accurate restoration of depressed articular surface could be achieved and the appropriate xation could be attained with no reduction loss. The isolated posterolateral tibia plateau fractures show a rare incidence among the proximal tibia fractures, and can be easily overlooked. However, the accurate reduction of articular surfaced and the restoration of lower extremity alignment are important to avoid the progression of traumatic arthritis, thereby can require more complex treatment. Inadequate reduction of articular surface and tibial alignment result in functional deterioration and persisted pain at involved limb.
To date, there have been several studies that segmented the tibial plateau fractures and proposed perspective surgical approach and xation techniques accordingly. Krause M et al. proposed the tencompartment classi cation system, divided the entire tibial plateau into anterior and posterior compartments at rst, then divided each anterior or posterior compartment into 5 sections [11]. However, their segmentation was just to describe the fracture con guration, not to provide clinical assistance for operation. Luo et al. proposed a three-column classi cation that classi es the tibial plateau into medial, lateral, and posterior as the center point between the two tibial spines, and explained that surgical planning can be made according to the classi cation [3]. However, they did not give more detailed explanations at posterolateral compartment.
Management of the tibial plateau fracture involving posterolateral compartment is technically challenging. There have been a few previous studies, suggested various surgical approaches and reduction techniques. Chih-Hsin et al. recommended the reduction and xation using the anterior approach for posterolateral tibial plateau fracture because the posterior approach and xation is risky to damage to the sural nerve, saphenous nerve, and popliteal arterial structures [12]. And they reported that satisfactory articular reduction and postoperative functional results can be achieved without postoperative complications even with reduction and xation using the anterior approach. On the other hand, Solomon et al. conducted a comparative study over the two-year follow-up period for the patients using the anterolateral approach and the patients using the trans-bular posterolateral approach. The results showed that the patients who underwent surgery by visually identifying the fracture site using the trans-bular posterolateral approach showed excellent postoperative results without formation of the articular step off in all patients compared to the patient group using the anterolateral approach with an average of 5.5 mm of large articular layer formation [13].
In this article, we proposed the sub-divided classi cation of the previous "three-column concept" [3] for the posterolateral tibial plateau fractures, and reported the clinical e cacy and safety of individualized approach and xation according the classi cation. The posterolateral area of the tibial plateau is restricted by the bula. Therefore, when surgeons make an attempt to access into most posterior area from anterior, they can encounter the di culty to identify with direct vision and get to have troublesome to achieve the accurate articular reduction. Also, when surgeons consider the posterolateral portion of the plateau as a column, they can experience problems at the exposure of fracture site and fracture manipulation by bular head and neurovascular structures around posterolateral knee.
We divided the posterolateral column to three zones based on bular head position, and suggested different surgical approachs, patients' position and xation methods accordingly. Our strategy was effective in establishing the precise planning of the surgical approach and xation methods. To apply the individualized approach, the entire tibial plateau should be identi ed in the axial or three-dimensionally reconstructed images of the preoperative CT scan. In the selected image, the anterior and posterior boundary of the bula, tibial spine, and the posterior sulcus of tibia can be indicated. The initial impacted area and extended fracture line can be identi ed and it can be classi ed according to the three-zone classi cation. Surgeon can make a surgical planning using one or more individualized approaches according to the corresponding zone, not to be afraid of using additional approaches.
In this study, all posterolateral tibial plateau fractures were treated based on the three-zone classi cation, and different approaches were not applied for each zone. Merely, in the third zone, the most posterior part, the posteromedial approach was used [14][15][16]. The posterolateral approach presents the possibility of iatrogenic neurovascular damage such as the anterior tibialis artery [17], and venous bleeding during soft tissue exfoliation for xation of distal screws often causes problems with hematoma and wound recovery [18]. Also, later, it is di cult to remove the plate due to the formation of scar tissue and adhesion on the plate when the plate needs to be removed later. Meanwhile, the posteromedial approach can secure a su cient surgical eld and supply e cient buttressing force from the posterior xation of the metal plate and screw. We kept to stand on that extended posteromedial approach would be more safe, easier and lesser burdensome than posterolateral approach to surgeon.
The postoperative results of the individualized approach to the posterolateral tibial plateau were excellent. In all patients, anatomical articular reduction was achieved and maintained without the reduction loss until the nal follow-up. Postoperative imaging showed a good alignment within the normal range. In all patients, satisfactory clinical outcomes and no complication was observed during the follow-up period.
The ndings of the present study are limited by its retrospective design with a limited number of samples. It was because that this study is designed only on the relatively rare isolated posterolateral tibial plateau fractures, which account for about 8% of patients with tibial plateau fractures. The surgical results can be in uenced by operator experience and would be have an inter-operator technique variation. The current study also lacks a control group and further studies on posterolateral plateau fractures of the tibia may be required to elicit clinical results for a larger patient group.

Conclusion
An individualized surgical approach and xation according to three-zone subdivision for isolated posterolateral tibial plateau fractures provided an effective and safe method to treat posterolateral tibial plateau fractures. Our strategy can be useful for reviewing the fracture pattern and con guration and establishing the preoperative planning of surgical procedures.     Radiologic evaluation to evaluate the anatomical reduction of the articular surface (a) Articular condition can be checked on simple radiograph taken after surgery. (b) The condition of articular reduction can be evaluated accurately on the coronal plane of CT scans