INTERNAL OSTEOSYNTHESIS OF DORSAL FRACTURES OF THE PROXIMAL TIBIA

Summary

INTRODUCTION: Fractures of the proximal tibia requiring treatment from dorsal approaches are uncommon. Nevertheless, they must be anatomically reduced and fixed in a stable manner. We use 3 surgical approaches to treat these fractures: posterolateral, posteromedial and posterior (direct posterocentral or modified posterocentral, as applicable). Individual approaches exist in various modifications. OBJECTIVE: Description of anatomical approaches to the dorsal portion of proximal tibia. To evaluate a cohort of patients treated with these surgical approaches with respect to CT findings of each fracture, choice of surgical approach, timing of surgery, type of fracture stabilization, peri- and postoperative complications, joint surface reduction and stabilization, and functional outcomes following each surgical approach and Lansinger score.

METHODOLOGY: A total of 26 patients (19 men and 7 women) who suffered proximal tibia fracture(s) requiring treatment from any of the dorsal approaches alone or in combination with any of the ventral approaches between January 2010 and December 2020 were included in the study. The average age of patients at the time of injury was 45.2 years (33-65 years) for men and 55.5 years (48-63 years) for women. Preoperative CT scan was performed in all patients.

In 14 cases there were avulsion fractures of the posterior cruciate ligament (PCL), in 12 cases dislocated fractures of the posterior column of the proximal tibia (these are further divided into isolated posterolateral (PL) fractures, isolated posteromedial (PM) fractures, and combined posterolateral/posteromedial fractures). Posterolateral approach was used in 4 patients, posterior – direct posterocentral – approach in 14 patients, posterior – modified posterocentral – approach in 4 patients and posteromedial approach in 4 patients. PCL avulsion injury was treated in all cases with 1-2 screws, supplemented with washers, if necessary. Isolated posteromedial fractures were treated with plate synthesis or with an individual screw. Isolated posterolateral fractures were treated with single screws. Combined fractures (posterolateral/posteromedial) required a combination of plate osteosynthesis and single screws.

RESULTS: The mean range of motion at 48 months was 0/0/130° in patients with isolated dorsal fractures, whereas in patients with combined fractures it was 0/5/120°. The average Lansinger score was 27 for avulsion PCL fractures, 26 for isolated PM fractures, 24 for isolated PL fractures and 22 for combined PM and PL fractures. Anatomical reduction of the articular surface was achieved in all (14) avulsion PCL injuries. Anatomical reduction was achieved in 3 cases (75%) of PM fractures and good reposition was achieved in 1 case (25 %). In the case of PL fractures, good reduction was achieved in 3 cases (75 %) and satisfactory reduction was achieved in 1 case (25 %). In all 4 cases of combined PM/PL fractures, anatomical reduction of PM fragment and good or satisfactory reduction of PL fragment were achieved (1 and 3 cases, respectively). Injuries to neurovascular structures, failure of osteosynthesis or deep infection after dorsal approaches were not reported.

CONCLUSION: Dorsal fractures of the proximal tibia need to be treated with one of the dorsal approaches. Each approach has its advantages and disadvantages, which every surgeon dealing with this issue should be informed about.

In the case of avulsion PCL fractures, the posterior approach in a form of direct posterocentral approach around the neurovascular bundle has proven reliable. In the case of isolated posteromedial fractures, the posteromedial approach is used. In case of isolated posterolateral fractures, the posterolateral approach is used. In the case of combined dorsal fractures (posteromedial/ posterolateral), on the other hand, a modified posterocentral approach completely outside the neurovascular bundle is recommended.

Keywords:

complications – Posterior tibial plateau fractures – posterior approaches – open reduction and internal fixation

OBJECTIVE

Description of anatomical approaches to the dorsal portion of proximal tibia. To evaluate a cohort of patients treated with these surgical approaches with respect to CT findings of each fracture, choice of surgical approach, timing of surgery, type of fracture stabilization, peri- and postoperative complications, joint surface reduction and stabilization, and functional outcomes following each surgical approach and Lansinger score.

TYPE OF STUDY: Retrospective study.

INTRODUCTION

Fractures of the proximal tibia are often complicated fractures of various spectre. There are many classifications of proximal tibia fractures. Each of them has both advantages and disadvantages.

The best known classifications of proximal tibia fractures are the AO classification (segment 41, types A1-C3), Schatzker classification (types I-VI) [45], and Moore classification [36, 37]. All the above-mentioned classifications take into account two-dimensional anteroposterior morphology (AO, Schatzker, Moore) or periarticular soft tissue injuries (Moore). A detailed description of these classifications is beyond the scope of this paper. Neither AO, Schatzker or Moore classification of proximal tibia fractures account adequately for posterior tibial plateau injuries. It was Luo [33] and subsequently Wang [54], who came up with the concept of 3-column fractures of the proximal tibia and the resulting threecolumn stabilization of these fractures on the basis of CT examination. This classification divides the proximal tibia into lateral, medial, and posterior columns, with the posterior column further subdivided into posterolateral and posteromedial portions. A column is considered fractured if its outer cortical bone is broken. A mere depression of the articular surface (i.e. a B2 fracture according to the AO) is not a column fracture. Fractures of one column are considered simple, fractures of two and three columns are considered complex.

This classification is also a guide to the treatment of individual columns. The 4-quadrant or 10-segment Krause classification used to plan the given surgical approach to treat individual injured segments is worth mentioning too [28].

Proximal tibia fractures are caused either by axial force (impaction or „split“ fractures) or rotational force (dislocation fractures) in a specific knee joint position (flexion, extension, varosity, valgosity).

In the case of monocondylar fractures (AO type B fractures), B2 and B3 fractures of the lateral tibial plateau predominate. Bicondylar fractures of the proximal tibia (AO type C fractures), on the contrary, usually have a posterolateral comminution and a posteromedial split fracture („high energy“ injuries), or a posteromedial comminution („low energy“ injuries) [28]. Low-energy injuries are typical especially in women of older age.

The history of posterior approaches to the proximal tibia dates back to 1945, when Abbott and Carpenter [1] described a posterior approach to the knee joint with partial dissection of the beginning of the medial head of m. gastrocnemius (hereafter MG), which was subsequently retracted laterally. This approach was then widely used for the surgical treatment of avulsion injuries of the posterior cruciate ligament.

The standard surgical approaches to the treatment of posterolateral or posteromedial fractures of the proximal tibia today are the standard posterolateral or posteromedial approaches according to Lobenhoffer [32], in which the head of the MG is abducted in the direction of the head opposite to this muscle, without the abducted head having to be dissected or longitudinally split. Many modifications have been described within posterolateral approaches – with/without fibular neck osteotomy, lateral epicondylar osteotomy [29], modified Frosch approach [12, 14]. The posteromedial approach can be used in either a prone or supine position.

There are also non-traditional procedures, such as medial approach with subsequent (via the ventral fracture line) elevation of the posterolateral impacted fragment, or the anterolateral approach with a bone window through which the posterolateral fragment is elevated again.

The so-called bilateral „dual plating“ - from the anterolateral and posteromedial approach – today it already represents the standard in the treatment of proximal tibia fractures. However, it needs be added that there are fractures – typically avulsion PCL injuries – that are difficult to treat from other approaches or are completely untreatable from the given approach.

From standard lateral approaches, it is usually not possible to anatomically reduce the posterior column fracture(s), let alone attach a buttress plate or possibly refix the PCL avulsion.

The operative principle is the refixation of avulsion PCL injuries and reconstruction of dorsal fractures of the proximal tibia. The aim of treatment of avulsion PCL injuries is to restore a stable and fully loadable knee joint. In the case of dorsal fractures of proximal tibia, the aim is to restore the anatomical „support“ of the dorsal portion of proximal tibia.

METHODOLOGY

The aim of the present study is to evaluate the elevenyear results of treatment of dorsal fractures of the proximal tibia treated at the authors‘ institution. A total of 26 patients (19 males and 7 females) were included in the evaluation, who suffered proximal tibia fracture(s) that were subsequently treated from one of the posterior approaches between January 1, 2010 and December 31, 2020; several patients also required treatment from the ventral approach at the same period or in the second period, as applicable.

The average age of patients at the time of injury was 45.2 years (33–65 years) for men and 55.5 years (48– 63 years) for women. The average length of patient monitoring was 48 months. Monotrauma was the exclusive mechanism of injury.

According to imaging scans, there were 2 types of proximal tibia fractures – avulsion PCL fractures and proximal tibia fractures with dorsolateral or dorsomedial fragment requiring posterior surgical approaches.

The clinical examination was followed by a basic imaging examination, i.e. X-ray of the knee in 2 mutually perpendicular projections. Furthermore, CT scans were added in all cases. In two cases we performed preoperative MRI scans. Preoperatively, we fixed the injured limb with a rigid brace.

Furthermore, we properly informed the patient about the course of the operation and the subsequent postoperative period (rehabilitation, or surgery in phase two).

We always operated under general anaesthesia, in the prone position, or on the side, as applicable, on a radiographically transparent table, with the tibia lightly supported, in a protected coagulum (Vulmizolin 2-3g i.v.; in case of allergy to betalactams - Dalacin 600mg i.v., possibly with Gentamycin 240mg i.v.). Further procedure varied according to the surgical approach used.

In the case of avulsion PCL injury (14 cases, see Table 2, Fig. 1), we used a posterior approach in the form of a posterocentral approach. Under aseptic precautions, we operated in the prone position with the knee slightly flexed; the tibia was supported by a foam pad. The „lazy S” character incision was made. However, inverted „L”- shaped incision is also described in the literature. In order to achieve a maximally atraumatic operation, we guided the incision in a sufficient length – about 12 cm. The skin incision was followed by dissection in the soft tissues with investigation of v. sapnena magna and n. suralis. The fascia between the two MG heads was cut longitudinally, followed by gentle blunt dissection around the neurovascular bundle. The bundle was subsequently retracted laterally together with the MG head. We did not perform dissection of MG medial head according to Trickey [53], described in the literature and widely used.

This was followed by vertical arthrotomy with utmost care to the course of PCL. After arthrotomy and subsequent evacuation of the hematoma, we mapped out the avulsion PCL fracture. Subsequently, we reduced the fracture with temporal stabilization using two K-wires under X-ray control, followed by successive drilling of 2 cannulated 4.0 cancellous screws, usually 55-60mm long. In 4 cases we used only one screw with a spiked washer.

In the case of isolated posteromedial fractures of the proximal tibia (4 cases, see Table 2, Fig. 2), the classic posteromedial approach in the prone position was used. The incision was made in the posteromedial portion of proximal tibia. After cutting through the subcutaneous tissue in the same extent and protecting the n. saphenus, fascia dissection followed. The common hamstring tendon was retracted ventromedially and the medial head of MG dorsolaterally, followed by pulling the m. popliteus apart. Dissection of the joint capsule was performed between the superficial portion of the LCM and the dorsal POL (“posterior oblique ligament”). After mapping out the fracture, it was stabilized.

In the case of isolated posterolateral fractures we used the posterolateral approach (4 cases, see Table 2, Fig. 3). We did the surgery in the prone position. In 2 patients we used posterolateral approach with fibular osteotomy. The skin and subcutaneous tissue incision was followed by dissection of n. fibularis communis up to the area of the fibular head. This was followed by osteotomy of the neck of the fibula and its retracting (tipping off) proximally, dissection of the joint capsule, exposure of the fracture and subsequent reduction of the fragments with their stabilization. Finally, stabilization of the osteotomy with a plate and screws. In the remaining 2 patients we used the posterolateral approach according to Lobenhoffer. After longitudinal skin incision and dissection of n. fibularis communis, we retracted the lateral head of MG medially, as well as the underlying tendon of m. popliteus. After horizontal capsulotomy under the lateral meniscus, we reduced and then stabilized the fractured fragment under visual control.

In the case of combined (posterolateral and posteromedial) fractures of the proximal tibia (4 cases, see Table 2, Fig. 4), we again used a posterior approach, this time in the form of a „modified“ posterocentral approach, which is performed completely outside the bundle. After intersection of the skin („lazy S“) and subcutaneous tissue, fascia dissection, retracting the medial head of MG laterally, gentle subperiosteal release of the course of m. soleus and m. popliteus, and subsequent insertion of Hohman elevators beyond the lateral margin of the proximal tibia, we treated the posteromedial fracture first using the plate osteosynthesis. There was always a solid bone fragment. The posterolateral fracture was treated after retraction of n. fibularis communis ventrolaterally and the lateral head of the MG medially. Thus we avoided any dissection around the neurovascular bundle. The fractures in the dorsal portion of the proximal tibia were reduced by ligamentotaxis in a slight extension to the maximal anatomical position; small bone chips were removed. In the case of central impressed fragments – typically in PL fractures – these fragments were elevated to the anatomical position (after the PL fragment had been tipped off) and the PL fragment was tipped back of them. In a large bone defect, allogeneic bone substitute was used in one case. This was followed by temporal stabilization with K-wires and, after radiographic control, plating (preformed plates on the proximal tibia, or „T“ plates on the distal radius, as applicable) with the drilling of individual screws. In the indicated cases, stabilization of PL fragments was performed with individual cancellous screws. Proximal screws in both posterior fragments must not be too long not to prevent the eventual reduction and fixation of the anterior column(s) in second stage. Treatment of anterolateral or anteromedial fractures does not allow this surgical approach, so these fractures were left – after regression of the swelling – for definitive treatment in phase two. In 2 cases, the soft tissue condition allowed for ventral ORIF immediately after dorsal stabilization. In the case of a severely comminuted fracture that required stabilization with external fixator in the first stage, we proceeded to plate osteosynthesis (from the posterior approach) after 19 days when the swelling had subsided; ORIF from the ventral approach was performed in the same patient 13 days after the previous procedure.

For surgical treatment of anterolateral or anteromedial fractures, repositioning of the patient is necessary. In the case of a two-phase procedure, there is also the option of a postoperative CT scan to review the previous surgery and also to plan treatment from a ventral surgical approach.

During hospitalization, thromboembolic disease was prevented (LMWH Clexane 0.4 ml or 0.6 ml once a day subcutaneously), sufficient pre- and postoperative analgesia, icing, ATB administered until the time of drain removal (usually after 24 hours – i.e. on the first postoperative day).

Postoperatively, patients were early verticalized with crutches. In case of an avulsion PCL injury, verticalization with crutches and a rigid brace (stepping with about one half of the body weight) followed, including rehabilitation on a motor plate without limitation. In the case of fractures of the proximal tibia, verticalization with crutches followed as well with stepping with about one quarter of the body weight; we did not apply a rigid brace. In both cases, we allowed full stepping after about 3-4 months.

We focused on the evaluation of CT findings in individual fractures, choice of surgical approach resulting from the CT, timing of surgery, type of stabilization, periand postoperative complications, evaluation of reduction and stabilization of the articular surface, functional results after individual surgical approaches, and Lansinger evaluation.

RESULTS

Evaluation of CT finding in each fracture (see Tab. 1)

According to the performed CT scan, the PL fractures were always impaction fractures with a split (AO type B3.1) requiring the tipping off of the dorsal fragment, elevation of the impressed fragment with potential allogeneic spongioplasty, subsequent tipping the dorsal fragment back with its temporary stabilization with Kwires, finally insertion of a plate or individual screws.

In the case of PM fracture, there was always a „solid“ bone block (AO type B1.3). In the case of avulsion PCL fractures there was always a bone fragment (AO type A1.3) of various sizes. In case of combined fractures (AO type C) there was always a combination of posterolateral impaction fracture and posteromedial split fracture.

In one patient with PCL avulsion, a fracture of the femoral diaphysis and a fracture of the proximal tibia requiring ORIF from the ventral approach were present, inter alia. In one patient, the avulsion PCL injury was initially managed conservatively. After 5 months, an MRI was performed with the finding of minimal healing progression, so surgical stabilization was proceeded to. In the same patient, arthroscopy was performed in phase two (within the interval of less than 3 months) due to a lateral meniscus tear on MRI.

Evaluation of the choice of surgical approach (see Tab. 2)

Posterolateral approach was used in 4 patients, posterior – direct posterocentral – approach in 14 patients, posterior – modified posterocentral – approach in 4 patients and posteromedial approach in 4 patients.

Evaluation of the surgery timing

The time from injury to definitive treatment averaged 5.7 days (1–19 days). Subsequent second surgery, if indicated (1x arthroscopy, 3x ventral ORIF, 1x dorsal ORIF), followed on average in 14.6 days after the injury (0–17 days).

Evaluation of the type of fracture stabilization

Individual fragments were stabilized with plates or individual screws. PM fragments were stabilized with a plate or single screw in isolated PM fractures. In the case of combined fractures, a plate was always used.

PL fragments in isolated PL fractures were stabilized with single screws or a plate. In the case of combined fractures, single screws or a plate were used.

Ventral fragments were stabilized with a plate. Ventral ORIF (AL and AM fragment) was indicated in a total of 7 patients (one with the avulsion PCL injury). In 4 cases, the ventral fragments were already stabilized during the first surgery; in the remaining 3 cases, the ventral fragments were stabilized in phase two. The average time from dorsal ORIF was 7.6 days (0–13 days). In one case (PM fracture), arthroscopic LCA reconstruction followed in an interval of 17 days.

Evaluation of perioperative and postoperative complications (see Tab. 3)

Postoperative complications (POC) were divided into 2 groups. Early POCs involved systemic complications, impaired healing, early failures of metallic material and nerve irritation, while late POCs included post-traumatic arthrosis and the need for another surgery.

As a part of early postoperative complications, we reported 2 hemodynamically insignificant pulmonary embolisms (despite adequate prevention of thromboembolic disease). In one case, the patient had, inter alia, a double segmental femoral fracture (treated with a retrograde nail in phase one) and a fracture of the anterolateral portion of the proximal tibia (treated in phase two); both fractures were on the left side, i.e. on the same side as the avulsion PCL fracture treated in phase three.

Infection requiring debridement and subsequent extraction of metallic material was observed once in a patient who also had ventral ORIF in addition to dorsal stabilization. We did not observe infection after dorsal ORIF in any of the above-mentioned approaches

Failure of OS material after dorsal ORIF did not occur in any case.

There was one need to „correct“ the metallic material. It was stabilization of an isolated PM fragment with a plate, where 2 of the screws were slightly protruding above the level of the ventral tibia, which required their shortening.

In the case of treatment of avulsion PCL injury, there was a slight elevation of one of the screws in one case; in subsequent X-ray examinations, the position of the screw was stationary and therefore its extraction was not indicated. We did not observe any nerve irritation.

In the case of avulsion PCL injuries, all patients were completely satisfied, capable of normal daily activities.

In the case of the above-mentioned comminuted fracture requiring temporary stabilization with an external fixator first and then stabilization from both posterior and anterior approaches, we performed extraction of all metallic material after 14 months.

As a part of late postoperative complications, we were forced to implant the total knee replacement in one patient 20 months after the injury due to severe posttraumatic gonarthrosis.

One patient had the arthtroscopy after 8 years and 2 months, the metal is still left in situ.

Extraction of metallic material in avulsion PCL injuries is a question. Due to the high risk of injury to neurovascular structures, extraction is not performed as a routine.

The same applies to dorsal fractures treated with plate osteosynthesis. Patients took an average of 8.5 months to return to work.

Evaluation of articular surface reduction and stabilization

Standard postoperative radiographs were used to assess the quality of joint surface reduction. No other methods of perioperative (arthroscopic) or postoperative (CT) control were used. Reduction with the articular shortfall below 1 mm was defined as an anatomic reduction, and reduction with the articular shortfall of 1-2 mm was defined a good reduction, and reduction with the articular shortfall over 2 mm was defined as a satisfactory reduction.

Anatomical reduction of the articular surface was achieved in all (14) avulsion PCL injuries. Anatomical reduction was achieved in 3 cases (75 %) of PM fractures and good reposition was achieved in 1 case (25 %). In the case of PL fractures, which - compared to PM fractures - included a comminuted zone, good reduction was achieved in 3 cases (75 %) and satisfactory reduction was achieved in 1 case (25 %). In the case of combined PL/PM fractures, the results were similar to those in isolated dorsal fractures. In all 4 cases involving a solid bony PM fragment, anatomical reduction was achieved; the results of PL fractures were inferior to those in isolated PL fractures, with good reduction achieved in 1 case (25 %) and satisfactory reduction in the remaining 3 cases (75 %).

Evaluation of functional results (see Tab. 1)

The average monitoring was 48 months (18–72 months). After this period, we assessed the overall range of motion and subjective patient satisfaction. Patients with avulsion PCL injury and patients with isolated posteromedial fractures achieved full range of motion after this time. In the case of isolated posterolateral fractures, the extension shortfall was 5 degrees, and flexion was not significantly affected. This slight limitation of extension is usually due to a greater tendency to scarring after the posterolateral approach. For combined dorsal tibial fractures, patients achieved an average range of motion of 0/10/125. Grade 4 post-traumatic arthrosis occurred in one case of combined fractures requiring extraction of the metallic material followed by total knee replacement. The average Lansinger score evaluating subjective satisfaction of patients after proximal tibia fractures was 27 for avulsion PCL fractures, 26 for isolated PM fractures, 24 for isolated PL fractures and 22 for combined PM/PL fractures. The score value correlated closely with the postoperative radiographic findings, where anatomical reduction was virtually achieved in avulsion PCL fractures, while good to satisfactory reduction was achieved in comminuted PL/PM fractures.

DISCUSSION

As early as 1945, Abbott and Carpenter described a direct posterior approach to the knee joint via the popliteal fossa [1]. In their paper they mention the dissection between the MG heads, i.e. around the neurovascular bundle in the popliteal region. In contrast, Burks wrote in his article: “The posterior approach to the knee is frequently viewed as difficult or hazardous and is therefore avoided when possible”.

Alpert, who gave a very detailed anatomical account of the direct posterior approach to the knee joint, writes in his paper: “However, use of this approach is uncommon and as a result, orthopedic residents and practicing orthopedic surgeons may not be familiar with the appropriate surgical anatomy” [2].

Most of the above-mentioned surgical approaches mobilise the medial head of the MG laterally or the lateral head of the MG medially, thereby completely avoiding any dissection around the bundle.

In the literature we find a disunion not only in the incidence of dorsal fractures of the proximal tibia, but also in the choice of surgical approach to these relatively complicated fractures.

The opinion on the choice of surgical approach to dorsal fractures of the proximal tibia is not uniform and undergoes a complex „evolution“ [1, 2, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 29, 30, 31, 32, 33, 34, 37, 37, 39, 40, 41, 42, 43, 44, 45, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57].

Lobenhoffer in 1997 was the first to come up with a posterolateral and posteromedial approach to the proximal tibia [32]. The posterolateral approach, which includes, in addition to multiplanar release of posterolateral joint structures (lig. meniscotibiale, posterolateral portion of the joint capsule, lig. popliteum obliuum, lig. popliteum arcuatum, tendon m. popliteus and external collateral ligament) also the obligatory OT of the proximal fibula (in the absence of a fracture of the proximal fibula), uses it to treat impression fractures of the posterolateral portion of the tibial plateau.

He also introduces the posteromedial approach, which he uses to treat dislocated fractures of the posteromedial portion of the tibial plateau, typically Moore I. The posteromedial portion of the proximal tibia is accessed by an incision between the internal collateral ligament and the oblique posterior ligament (“posterior oblique ligament”). If necessary, the dissection of pes anserinus is performed. It should be noted that both of these approaches are performed in the supine position. In our opinion, this position of the patient with a flexed knee not only makes it difficult to reduce exactly the PM fragment, which is most distally dislocated in knee flexion, but also to place correctly the plate, which should be placed sufficiently „dorsally“ to achieve compression perpendicular to the fracture line. Lobenhoffer himself used the combined PM and PL approach „only“ in 3 cases.

Frosch [12, 13] modifies the posterolateral approach described by Lobenhoffer. From one posterolateral incision, he performs both lateral arthrotomy (ventrally from the biceps tendon), through which he subsequently inspects the articular surface and at the same time visually checks the reduction (which is hardly possible by the arthrotomy alone), and the actual reduction and stabilization of the fracture (dorsally from the biceps tendon between the lateral head of the MG and m. soleus, then between m. popliteus and m. soleus). In his paper, he explicitly writes: “Die gute Darstellung der posterolateralen Ecke führt aber zu einem nicht unerheblichen Weichteiltrauma“. They try to minimize the devastation of soft tissues (see above), fibular OT is not performed as standard.

As for the actual posterior approaches to the proximal tibia, there are practically 3 alternatives. The so-called direct posterocentral approach around the bundle, i.e. between the two MG heads [17, 38, 39], the so-called modified (limited) posterocentral approach going medially from the medial MG head or laterally from the lateral MG head, i.e. completely outside the bundle [7, 8, 11, 25, 26, 33] and finally the so-called „direct“ posterior approach with an incision above the medial MG head [15, 16]. Other approaches are derived from them or use only a part of them.

Carlson [7, 8] uses a „dual“ incision in the form of 2 separate „lazy-S“ dorsal approaches in the treatment of dorsal tibial plateau fractures. His quotation is very interesting: “Why is it advantageous to expose both posterior condyles from separate incisions? A single S-shaped incision will limit the distal exposure of one of the condyles. With the elevation of the large skin flap necessary for visualization, injury to the sural nerve is much more likely. Thus, to gain adequate exposure to buttress plate both posterior plateaus, the dual incision approach is felt to be best.”

Practical „opposite“ is presented by Muhm [38, 39], who uses just one – so-called posterocentral – approach around the bundle [ 38, 39] to treat both avulsion PCL fractures and dorsally dislocated fractures of the proximal tibia. His paper emphasizes careful anatomy of the popliteal region. This approach is not burdened by any significant „mobilisation“ or soft tissue release, as in the PL approach (OT of the proximal fibula, release of the lateral ligamentous stabilizers - meniscotibial ligament, PL portion of the joint capsule up to the attachment of the LP obliquum and arcuatum) or the PM approach (release of the medial capsuloligamentous stabilizers). The same method, i.e. dissection around the bundle, is described by Gardner as well [17].

Hsieh [23] on the other hand, creates a cortical bone „window“ at the site of depression in the AL approach, then reduces the posterolateral fracture, inserts an autologous bone graft into the resulting effect, and then stabilizes the fracture with a plate with posterolaterally directed screws.

Chen [25, 26] uses – in both PL and PM fractures – „posterolateral“ and „posteromedial“ inverted L-shaped approaches, with the horizontal shoulder located approximately 3 cm proximal to the flexor popliteal groove. The quotation marks are here because – although the dissection is the same in both approaches as in the standard PL and PM approaches – the given approaches are „more“ dorsal on both sides of the MG. In the case of bilateral fractures of the dorsal portion of proximal tibia, the authors use a combination of both above-mentioned approaches, first treating the PM split fracture and then the PL compression fracture.

Potocnik [42] performs reduction of an impacted PL fragment from the medial approach in a relatively unusual manner, elevating it anteromedially from the main fracture line to a satisfactory position under radiographic control and then percutaneously stabilizing it with screws from the lateral mini-incisions. It is only after stabilization of this regment that the PM fragment is stabilized from the same incision. This procedure is opposed by Gosling [19], who mentions that laterally placed screws or plate cannot sufficiently stabilize the PL fragment in case of bicondylar fractures.

Luo [33] even treats the entire dorsal column (with the medial column, if appropriate) from a single dorsal medial inverted „L“-shaped incision (so-called „posterior reversed L-shaped approach“). From one incision he is able not only to stabilize the PM fragment, but also – after retracting the medial head of the MG laterally – to treat and stabilize the PL fracture.

However, it should be noted that the posterolateral fragment is stabilized by a placed that is placed „obliquely“ compared to the vertically oriented medial plate. He also introduces into terminology the concept of „floating position“, in which the patient is in the position on the side. In this position, fractures of all three columns of the tibial plateau can be treated simultaneously without postoperative rotation of the patient. In the case of a lateral column fracture, he performs its stabilization from a separate anterolateral approach. His quote regarding the posterolateral approach is interesting: “In the authors’ opinion, this approach also obviates the use of a second posterolateral incision as advocated by Carlson. The posterolateral approach creates problems around the exposure of the common peroneal nerve and management of posterior tibial recurrent artery (a branch from the proximal portion of the anterior tibial artery). Through the inverted L-shaped posterior approach, the reduction of the posterolateral articular surface can be achieved with an elevator through a posterior ‘‘fracture window.’ “

The same method, i.e. inverted „L“-shaped incision followed by retracting the medial head of the MG laterally, is described e.g. by Faucett [11].

However, it seems to us that „dragging“ the medial head of MG laterally may not be sufficient and may contribute to reduced visualization of the PL fragment and its resulting lack of stabilization.

Of the same opinion is also Orapiriyakul [40], who recommends performing a separate posterolateral approach when the fracture is within the range of 44-81% of the width of the lateral tibial plateau. According to him, this so-called „blind area“ is very difficult to treat from the posterior approach according to Luo.

Galla [15, 16] comes up with a „direct“ posterior approach to the proximal tibia via the medial head of the MG, which he then bluntly mobilizes and retracts laterally along with the neurovascular bundle situated laterally from the MG head. Thus, the posteromedial portion of the posterior column of the proximal tibia is gracefully accessed without opening the joint cavity or significant traumatisation of the soft tissues (MG dissection or release of the pes anserinus). The position of the patient on the abdomen facilitates the reduction of the dislocated PM fragment followed by the accurate application of the plate. Both of these can be more challenging in the case of the supine posteromedial approach according to Lobenhoffer. As already mentioned, the plate should be placed sufficiently dorsally so that the compression is perpendicular to the fracture line. This is, in our opinion, difficult to achieve from the posteromedial approach according to Lobenhoffer. In the case of lateral or bicondylar fractures of the proximal tibia, he either extends the original approach craniocaudally, or uses a separate posterolateral approach. In the case of the „extension“ approach, he bluntly mobilizes the lateral head of the MG, which he subsequently then retracts medially.

Many studies describe the incidence and morphology of posteromedial and posterolateral fragments.

Yang [56] reported the presence of a fracture of the dorsal portion of the proximal tibia in 151 of the total of 525 patients (28.8 %). In 69 patients (45.7 %) there was a posteromedial fragment, in 53 patients (35.1 %) there was a posterolateral fragment, and in 29 patients (19.2 %) there was a combination of both fragments.

Of 111 bicondylar fractures, Higgins [22] finds a posteromedial fragment in 65 cases (59 %)

Barei [3, 4] evaluates a group of 146 bicondylar fractures of the proximal tibia (i.e. AO type C).

The presence of the PM fragment is found in 28.8 %; the presence of PL fragments is not considered.

Sohn [46], in contrast, evaluates the presence of PL fragment in 190 proximal tibia fractures, finding it in 84 fractures (44.2 %), while it was present in 35.9 % of type B fractures (37 fractures out of a total of 103) and 54.2 % of type C fractures (47 fractures out of a total of 87).

Lobenhoffer [32] reported an incidence of isolated PM fractures (AO type B) of 6%, in the case of bicondylar fractures (AO type C) of 1 5%.

CONCLUSION

There are not many fractures of the proximal tibia requiring treatment from dorsal approaches. This is also evidenced by our patient spectrum, with only 26 patients treated in this way over 11 years. If we confront this number with the world literature, we get similar figures. Their definitive treatment should therefore be carried out in a specialised department by an erudite trauma surgeon proficient in this approach. Although lateral and medial stabilization („double plating“) of proximal tibia fractures represents very good clinical results and is the „gold standard” [10, 19, 57], many recent authors repeatedly mention the importance of dorsal stabilization of the proximal tibia [35].

This is due to the fact that „double plating“ can hardly stabilize the fractures of the dorsal tibia, which often take place in the coronal plane. Krause in his „10 segment classification“ concludes that dorsal fractures of the proximal tibia occur universally in type B and C fractures according to the AO classification [28].

A relatively low number of patients treated with these approaches is the disadvantage of the presented paper. This is because a large proportion of proximal tibia fractures is treatable from the ventral approaches.

An indisputable advantage of posterior approaches is their use even when the soft tissue situation is unfavourable ventrally.

Disadvantages include the dissection around neurovascular bundle (in case of direct posterocentral approach), larger skin incision and release (in case of plate application) of m. soleus or m. popliteus, as well as the necessity of ventral approaches in case of their indication.

Indications for these surgical approaches include avulsion PCL injuries and dislocated dorsolateral or dorsomedial fractures of the proximal tibia that would otherwise be difficult to treat or even untreatable by other approaches.

The paper is limited by a relatively small cohort of patients, by the fact that standard X-ray examination was used for postoperative follow-up instead of CT, and in considerable part by the fact that many proximal tibia fractures are treatable from other – ventral – approaches.