1 Introduction

Tibial plateau fractures are intra-articular fractures. They occur with high-energy trauma in young people and lower-energy trauma in elderly individuals.¹ Tibial plateau fractures account for 1% of all fractures.² High-energy trauma in young people is more difficult to treat. In these patients, accompanying soft tissue problems and intra-articular fragmentation may occur together. As these patients are younger, the performance requirements are higher. Therefore, all fracture fragments must be fixed anatomically and with sufficient stability. Otherwise, severe joint movement limitation, deformity, early arthrosis, and serious loss of function may occur.³

Open reduction and plating are frequently used in the treatment of tibial plateau fractures and satisfactory results are obtained.⁴ External fixation is preferred temporarily for patients with soft tissue problems.⁵ Plate fixation is a good option for both providing and maintaining anatomical reduction.⁶ Plating can be applied with lateral, medial, and posterior incisions in cases of tibial plateau fractures. Skin problems are particularly more common in the medial incision area.⁷

More recently, the importance of posteromedial fragments in cases of tibial plateau fractures has begun to be understood.^8,9 The posteromedial fragment is the fracture fragment in the posterior region when the medial tibial plateau is divided into two.^8,9 This has a significant impact on the stability of the knee joint biomechanically. When it is not properly detected, severe deformity, loss of reduction, limitation of joint motion, and early development of arthrosis may occur.^10,11

In this study, we investigate whether posteromedial fracture fragments influence the clinical results, radiological findings, and complications in the treatment of tibial plateau fractures. Our hypothesis is that posteromedial corner fragments affect the clinical and radiological results.

2 Material and methods

Patients who underwent plate fixation due to tibial plateau fractures in our clinic were identified through the hospital automation system after the research was approved by the relevant ethics committee (Ege University Clinical Ethics Committee, No: 70198063–50.06.04 16-10/10). Open fractures, pathological fractures, patients with insufficient radiological clinical data, patients with a clinical follow-up of less than 12 months, and pediatric patients were excluded from the study. Patients whose treatment was completed with conservative, screw-only, and external fixation methods were also excluded. According to the type of fracture, simple falls were considered as low-energy injuries, while traffic accidents, occupational accidents, falls from heights, sports injuries, and motorcycle accidents were considered low-energy injuries.

Fracture typing of the patients was performed according to the Schatzker classification.¹² Laterally undissociated fractures were considered as type 1, laterally separated fractures as type 2, laterally collapsed fractures as type 3, medial fractures as type 4, bicondylar fractures as type 5, and fractures extending to the tibial shaft as type 6. Preoperative evaluations of fractures were performed with well-directed knee and tibia radiographs and computed tomography (CT). In accordance with the musculoskeletal CT protocol, scanning was performed at 1-mm intervals and source images were obtained. These were evaluated with reformatted images at 2–3 mm in all three planes (axial, coronal, and sagittal). The presence of a posteromedial fracture fragment was evaluated according to the CT sections^8,9 (Fig. 1).

Close

Fig. 1 Posteromedial fragments in CT sections (al: anterolateral, pl: posterolateral, am: anteromedial, pm: posteromedial, white arrows point to posteromedial fragments).

Export to PPT

The timing of the fracture treatment was decided according to the condition of the soft tissue. Patients with minimal soft tissue edema were followed in the initial period, and patients with higher levels of soft tissue edema were first followed with an external fixator that went beyond the knee (Fig. 2). The second sessions were held after 3 weeks. Surgical procedures were performed under spinal/epidural anesthesia without the use of a tourniquet. Antibiotic prophylaxis with 1 g of cefazolin sodium was administered before the surgical procedure. A lateral anatomic locking plate was fixed to the lateral fracture fragment with a lateral incision, and a locking anatomical plate was fixed to the medial fragment with a medial/posterior incision. A double incision procedure was applied for patients with two colon fractures (medial + lateral). Compression screws were applied outside the plate to increase reduction and stabilization according to the fracture geometry. In the presence of a posteromedial fragment, a posteromedial/posterior incision was used and a posteromedial plate was applied (Fig. 3). Iliac wing bone grafting was applied for patients with bone defects.

Close

Fig. 2 Bicondylar tibial plateau fracture with high-energy posteromedial fragment. In the first stage, this patient was followed with external fixation until soft tissue healing was completed. Reduction of the intraoperative medial fragment; first-year postoperative control radiographs and knee movements.

Export to PPT

Close

Fig. 3 Fixation of tibial plateau fracture with posteromedial fragment with posteromedial plate and compression screw.

Export to PPT

After treatment, the patients were hospitalized for 1–2 days. After the operation, deep vein thrombosis prophylaxis was applied for 10 days with low-molecular-weight heparin (enoxaparin sodium) at 1 × 0.4 mL. Prophylactic treatment was continued for 30 days with 100 mg of acetylsalicylic acid at 1 × 1. Antibiotic prophylaxis with 1 g of cefazolin sodium at 3 × 1 was given for 48 h. For patients with wound discharge, antibiotics were continued according to the antibiogram culture. In the early post-treatment period, they were mobilized with touch down weight-bearing. Joint movements were started as actively as the patient's pain allowed. Long-leg adjustable-angle knee braces were used in the postoperative follow-up. The sutures were removed in the second week. Mobilization with 50% load was allowed after a check-up appointment in the first month. Full weight-bearing was started after 2 months. The patients were called in for regular check-ups at the 1st month, 3rd month, 6th month, and 1st year after the operation and clinical and radiological evaluations were made. Concomitant problems and complications were noted.

In the radiological evaluations, the anterior and lateral proximal tibial angles (APTA and LPTA) were measured early after the operation and during check-ups. The PACS system of the hospital's automation system was used for angle measurements. Angle measurements on direct radiographs were performed by an experienced orthopedic specialist. The validation of the obtained angle measurements was done by another experienced orthopedic specialist. The normal value was accepted as 90° for the APTA and 80° for the LPTA. Angle deviations of 5° and more than 5° were considered as evidence of malunion for these respective measurements.¹³ If this difference was observed in the first and following radiographs, it was considered as reduction loss. Joint separation was defined as separation of more than 2 mm on the joint surface after reduction. Absence of bone bridging for more than 6 months on follow-up radiographs was considered as nonunion. Arthrosis was defined as increased narrowing, sclerosis, and new osteophyte formation on control radiographs between the femoral and tibial joint surfaces compared to the first radiograph. Arthrofibrosis was defined as less than 90° of flexion and more than 10° of extension loss. Purulent discharge, osteolysis, debridement under operating room conditions, and removal of implants were accepted as signs of infection. Superficial infections that healed with simple local debridement, wound dressing, and antibiotic therapy were not considered as deep infections. Clinical and radiological evaluation was done according to the Rasmussen criteria.¹⁴ The Rasmussen scores of the patients were evaluated by a single experienced orthopedic surgeon. Clinical Rasmussen scores were classified as excellent (27–30), good (20–26), moderate (10–19), and poor (<10). Radiological Rasmussen scores were classified as excellent (18), good (12–17), moderate (6-11), and poor (<6).

The obtained patient data were uploaded to Microsoft Excel. Statistical analysis was done with SPSS. Since this study was not a prospective study, sample size and power analysis were not performed. The patients were divided into two groups as those with and without posteromedial fracture fragments. Age, gender, side, follow-up time, APTA, LPTA, clinical and radiological scores, and whether there were complications were evaluated statistically between the two groups. Chi-square tests were used to compare cross-sectional data. If the minimum value in the table was less than 5, Fisher's exact test was used, and if it was higher, Pearson's chi-square test was used. In the evaluation of numerical data, the Shapiro-Wilk test was used to determine whether the distribution was normal or not. Parametric tests (t-tests) were used when the distribution was normal and non-parametric tests (Mann-Whitney U, Wilcoxon's test) were used when it was not. The significance level was accepted as p < 0.05.

3 Results

Fifty-two patients were identified by scanning the clinical archive. Thirty-eight patients with adequate clinical and radiological follow-up were included in the study. Twenty-six of the patients were male and 12 were female. The mean age was 46.08 (24–80) years. Twenty-one had undergone an operation on the right side and 17 on the left side. The mean follow-up period was 18.55 (12–36) months (Table 1). Posteromedial fragments were present in 21 (55.6%) cases and absent in 17 (44.4%) cases.

Patients were classified according to Schatzker's system, with type 2 being seen in 6 cases, type 4 in 8 cases, type 5 in 14 cases, and type 6 in 9 cases. Six patients had been operated on for simple falls, 9 for traffic accidents (inside and outside of vehicles), 14 for motorcycle accidents, 6 for falls from a height, 2 for work accidents, and 1 for a sports injury. Seven patients had additional fractures (1 rib, 2 distal radius, 1 contralateral tibia, 1 metacarpal, 1 contralateral femur lateral condyle, and 1 forearm and elbow).

A single lateral locking plate was placed for 12 patients, a medial locking plate for 11 patients, and both medial and lateral plates for 15 patients. Due to soft tissue edema in 3 cases, external fixation was applied in the first session and then the plate was applied in the second session. Iliac wing bone grafting was applied for 22 patients.

The mean early postoperative APTA of the patients was 88.21° (85–90°) and the mean follow-up value was 87.86° (80–90°). The mean early and follow-up LPTA angles were 79.97° (73–87°) and 79.92° (74–89°), respectively. Angle differences of >5° for the APTA were observed in the early postoperative period for 2 patients and during follow-up for 5 patients. For the LPTA, there was an angle difference of more than 3° in the early postoperative period and of more than 5° during follow-up for 3 patients. Thirteen patients had separation of more than 3 mm. The mean Rasmussen clinical score of the patients was 27.92 (17–30) and the mean radiological score was 15.89^10–18 (Table 2).

Complications developed in 6 cases. For 1 patient, healing was achieved by loosening the locking screws from the plate. The impact on the clinical results was minimal. Infection and nonunion occurred together in 1 case (Fig. 4). Joint movement limitation and arthrofibrosis developed in 2 cases. Closed manipulation was applied. For 2 patients, osteolysis occurred at the fracture line on the lateral plateau but did not require additional surgery.

Close

Fig. 4 Fixation of a posteromedial fragment with posteromedial plate first. Failure due to infection was detected afterwards. External fixation and fixation of fracture fragments with screws. Early osteoarthritis development at 24 months after the operation.

Export to PPT

When the patients were divided into 2 groups as those with and without posteromedial fracture fragments, the statistical evaluation of the groups showed similar age, gender, and follow-up time distributions (p > 0.05). The right side was more often affected in patients with posteromedial fragments (p < 0.05). Graft use rates, additional fracture rates, and low-energy versus high-energy injury rates were similar between the two groups (p > 0.05). There was no difference in early postoperative and follow-up APTA and LPTA between the two groups (p > 0.05). When the patients with and without posteromedial fragments were examined in terms of changes in the early and follow-up APTA and LPTA in dependent groups, no significant change was observed (p > 0.05). Rasmussen clinical and radiological scores were also similar between the groups (p > 0.05).

In the group with posteromedial fragments, infection in 1 (4.8%) case, nonunion in the same case (4.8%), 7 (33.3%) cases of loss of reduction, 7 (33.3%) cases of malunion, 8 (38.1%) cases of joint separation, 2 (9.5%) cases of early arthrosis, and 2 (9.5%) cases of arthrofibrosis were observed. Infection, nonunion, loss of reduction, and arthrofibrosis were not observed in the group without posteromedial fragments. There were 2 (11.8%) cases of malunion, 5 (29.4%) cases of joint separation, and 1 (5.9%) case of early arthrosis. There were no statistically significant differences between the two groups in rates of infection, nonunion, malunion, joint separation, early arthrosis, and arthrofibrosis (p > 0.05). There was a statistically significant difference between the two groups in terms of reduction loss (p: 0.011, p < 0.05) (Table 3).

4 Discussion

Tibia plateau fractures are difficult to treat and prone to complications. In recent studies, the importance of fractures in which the posteromedial corner of the tibial plateau is affected has been emphasized.^15–17 Posteromedial fragments are seen in 29–59% of cases of plateau fractures.^8,9 This rate was 55.6% in our patients.

Good early results have been reported with open reduction and internal fixation in the treatment of tibial plateau fractures.^6,18 Ozkaya et al. in a study with a mean follow-up period of 26 months, did not encounter reduction or loss of alignment and early arthrosis in any of their patients and clinical scores were satisfactory.⁶ Biggi et al. obtained excellent results for 78% of the patients in a retrospective cohort study according to the Rasmussen criteria, with patients being followed for an average of 18 months.¹⁸ In our study, the mean follow-up period was 18.55 months. There was loss of reduction in 18.4% and early arthrosis in 7.9% of our cases and all patients with reduction loss had posteromedial fragments. Malalignment was present in 23.7% of our cases. According to the clinical Rasmussen criteria, we achieved excellent results at a rate of 81.6% (Tables 2 and 3).

Complications after internal fixation of tibial plateau fractures include joint stiffness (0–37.5%), infection (2–11%), late-stage arthrosis (23.8–44%), malunion (34%), and nonunion (4%).^19,20 Among our patients, we observed joint stiffness in 2 cases (5.3%) and infection in 1 case (2.6%). In addition to the infection, we also experienced 1 case of nonunion (2.6%) in the same patient. We faced malunion problems in 9 cases (23.7%). Our infection rate was lower than those reported in the literature, mostly likely because we preferred staged surgery for 3 patients with soft tissue problems, soft tissues were preserved using double incisions when necessary, and open fractures were not included in this study. Posttraumatic arthrosis was observed in only 3 (7.9%) of our patients but our follow-up period was short. This rate is likely to increase in later follow-up periods.

The understanding of the importance of posterior coronal fractures in tibial plateau fractures has started to increase.¹¹ Failure to recognize or consider a posterior sagittal fracture fragment in the treatment of tibial plateau fractures has been associated with poor clinical outcomes. Therefore, special fixation for the posterior fracture fragment, whether lateral or medial, has been recommended.¹¹ Weaver et al. found that in cases of tibial plateau fractures, when there was a posterior coronal fracture fragment in the medial part, there was a median loss of fixation of 2.0° in patients who underwent lateral plating. In patients without a posteromedial fragment and with lateral plating, this was 0.5°, and they found a statistically significant difference between the two groups.²¹ None of our patients with a posteromedial fragment were treated with a lateral plate alone. However, in our comparison of patients with and without posteromedial fragments, all patients with loss of reduction had posteromedial fragments (7 patients, 33.3%). There was no loss of reduction among our patients without posteromedial fragments. There was a statistically significant difference between the two groups in our study (p: 0.011, p < 0.05). Due to the anatomical structure of the posterior proximal tibia, these fractures are in a difficult region for dissection, plate-screw placement, and rigid fixation.¹⁰ However, poor fixation of the posterior part is associated with loss of reduction and poor clinical outcome. Therefore, when medial-lateral incisions are not sufficient in this region, fixation with posterior incisions is recommended as necessary.¹⁰

Our study has some limitations. First, the study design was retrospective and non-randomized. Randomization would have been difficult as we compared clinical outcomes according to fracture type, not differences in tibial plateau fracture treatment modalities. Second, the number of patients was small. Third, our follow-up period was short. A longer follow-up period would be crucial in accurately evaluating the development of arthrosis.

Posteromedial tibial plateau fractures are specific fractures. Anatomical and stable fixation of parts that are difficult to reach plays a key role in treatment in obtaining satisfactory results. Reasons for failure include non-anatomical reduction, infection, and arthrofibrosis. The presence of a posteromedial fragment is a key factor in reduction loss.

Device status statement

The manuscript submitted does not containing formation about medical device(s).

Ethical approval

Ethics approval for this study was obtained from the Ethics Committee of Ege University, Turkey.

Declaration of conflict of interest statement

The authors declare that they have no conflict of interest.

Funding

We affirm that we have no financial affiliation (including research funding) or involvement with any commercial organization that has a direct financial interest in any matter included in this manuscript.

Ethics committee approval

IRB no:16-10/10 (Ege University Ethics Committee).

Informed consent

Consent was not obtained because of the retrospective nature of the study.

CRediT authorship contribution statement

Mehmet Akdemir: Data curation, Writing – original draft, Formal analysis, Supervision. Mehmet Aykut Türken: Visualization, Investigation, Resources. Ahmet Cemil Turan: Conceptualization, Methodology, Software. Ahmet Çağdaş Biçen: Software, Writing – review & editing. Ali İhsan Kılıç: Conceptualization, Validation, Data curation, Resources.