Introduction

Good clinical results have been reported with open-wedge high tibial osteotomy (OWHTO) without bone grafting with locking plates that act as internal fixators [2,3,4, 7, 15, 19]. However, despite good clinical outcomes, complications, such as delayed bone healing and nonunion, have been reported [21, 26]. Although there are some reports regarding risk factors of delayed bone healing, such as obesity, smoking, and lateral hinge fractures [11, 18], few studies have evaluated in detail the progression of bone formation in the osteotomy gap after OWHTO. A retrospective study by Staubli et al. [23] of 53 cases of OWHTO with TomoFix® plate (DePuy Synthes, Solothlun, Switzerland) without filling the gap showed that bone healing of the defect progressed from the lateral hinge to the medial side and that 75% of the gap was filled with new bone within 6–18 months on the basis of standard radiographic assessment. Brosset et al. [1] prospectively performed 51 OWHTOs with TomoFix plates without filling the gap and reported that radiological bone union and filling of 4/5 of the osteotomy site was obtained in 49 (96%) of 51 knees after an average of 4.5 months. However, the definition of gap filling on anteroposterior (AP) radiographs was not clear in their reports. It is crucial for the evaluation of bone formation after OWHTO to clearly define bone formation at the osteotomy site. Furthermore, to date, the progression of bone formation after plate removal has not been investigated, and there is a lack of information regarding the optimal timing for plate removal after OWHTO.

The purposes of this study were (1) to clearly define the bone formation at the osteotomy site; (2) to evaluate the progression of bone formation at the osteotomy gap after OWHTO, including after plate removal; and (3) to investigate risk factors for delayed bone healing. It was hypothesized that gap filling progresses from lateral to medial after OWHTO without bone grafting and progresses further after plate removal and that unstable lateral hinge fractures would cause delayed gap filling after OWHTO.

Materials and methods

This retrospective case series was approved by the Ethics Committee of Toyama Municipal Hospital (IRB No. 2014-14), and all patients provided informed consent prior to participating in the study. Ninety-three patients (102 knees) with OA or osteonecrosis (ON) who underwent OWHTO without bone grafting between February 2007 and August 2015 were included in this study. There were 80 knees with medial knee OA and 22 knees with ON of the medial femoral condyle. The patients’ mean age was 62.6 ± 11.2 years (range 22–83 years), and the mean body mass index (BMI) was 24.4 ± 3.0 kg/m2 (range 16.1–31.3 kg/m2) at the time of surgery. Patients with complete postoperative follow-up records for ≥ 2 years were included, and the mean follow-up period was 62.3 ± 30.2 months (range 25–126 months). All patients had the plate removed at > 1 year after surgery, and the mean interval from surgery to plate removal was 17.6 ± 5.1 months (range, 12–38 months). The patient characteristics are shown in Table 1.

Table 1 Patient demographic data
Full size table

Our inclusion criteria for the OWHTO procedure were as follows: (1) symptomatic medial OA or ON of the medial femoral condyle, (2) varus malalignment, and (3) active patients with good postoperative rehabilitation program compliance. There were no age restrictions. The contraindications for OWHTO were (1) history of joint infection, (2) symptomatic OA of the lateral compartment or patellofemoral joint, (3) joint instability, (4) femorotibial angle (FTA) > 185°, and (5) flexion contracture > 15°.

Surgical procedure and postoperative rehabilitation

An AP long-leg weight-bearing radiograph was used for preoperative planning. The weight-bearing line was aimed at a point 65–70% lateral on the transverse diameter of the tibial plateau. Arthroscopy was routinely performed prior to HTO to evaluate the medial, lateral, and patellofemoral cartilage. The medial proximal tibia was exposed by an oblique incision, and the superficial fibers of the medial collateral ligament were released distally. The upper border of the pes anserinus was defined as the starting point of the osteotomy. Two K-wires directed just proximal to the tibiofibular joint were used as place markers for the saw cut. The lateral-most 10 mm of the tibial head was left intact and used as a hinge for the osteotomy. The separate ascending cut of the biplanar osteotomy was made 1.5 cm behind the tibial tuberosity in the frontal plane at an angle of 100°–110° to the first osteotomy plane. Several chisels were used to gradually open the osteotomy, and a laminar spreader was used for final opening. The TomoFix plate was inserted into a subcutaneous tunnel formed on the medial side of the tibia and fixed in place with eight locking screws. No bone graft or bone substitute was placed in the osteotomy site.

Isometric quadriceps, active ankle exercises, and straight leg raises were started on the first postoperative day. Partial weight-bearing started 2 weeks after surgery in the initial 52 knees and 1 week after surgery in the remaining 51 knees. Full weight-bearing was permitted after 4 weeks.

Radiological evaluations

Radiological evaluations were performed at the following time points: preoperative consultation; 3, 6, and 12 months postoperatively; at plate removal; and at last follow-up (mean 62.3 ± 30.2 months). The weight-bearing line ratio (WBLR) and FTA were evaluated to assess alignment. WBLR was measured from standing AP whole-leg radiographs. To calculate WBLR, a line was drawn from the center of the femoral head to the midpoint of the proximal talar joint surface. WBLR was defined as the horizontal distance from the WBL to the medial edge of the tibial plateau, divided by the width of the tibial plateau. A standing AP view, with an extended knee joint, was used to assess FTA, and FTA was defined as the lateral angle between the axis of the femoral and tibial shaft.

Radiographic evaluation of bone formation in the osteotomy gap

Figure 1 shows an AP radiograph taken at plate removal. It is difficult to determine to what extent bone formation progressed, zone A or zone B. The definition of bone formation is important when evaluating bone formation in the osteotomy gap. Therefore, first, we divided the osteotomy gap into the lateral hinge and the four zones on AP radiographs (Fig. 2). Because it is difficult to determine the extent of bone formation, we evaluated to what extent gap filling progressed. To confirm this, we compared bone formation in the osteotomy gap on AP radiographs and sagittal multiplanar reconstruction computed tomography (MPR-CT) performed after plate removal in the first 60 knees. Two independent observers (K.G. and A.N.) evaluated this twice, with a 6 week interval between analyses. Both observers were blinded to the results of previous observations. As a result, the zone in which gap filling is observed on sagittal MPR-CT is the same as the zone in which trabecular bone continuity can be observed on AP radiographs (Fig. 2). Inter- and intra-observer reliability of gap filling measurements on AP radiographs, quantified by the kappa coefficients, were 0.770–0.820 and 0.778–0.884, respectively. In addition, the kappa coefficient to determine the inter-observer agreement for diagnosis between radiographs and CT was 0.770–0.885. Therefore, gap filling was defined as the zone in which trabecular bone continuity can be observed on AP radiographs. According to this definition, we evaluated bone formation in the osteotomy gap at 3, 6, and 12 months postoperatively; at plate removal; and at the final follow-up.

Fig. 1
figure 1

Anteroposterior radiograph taken at plate removal. It is difficult to determine to what extent bone formation progressed, zone A or zone B

Full size image
Fig. 2
figure 2

a Anteroposterior radiograph showing the evaluation method for bone formation. The osteotomy gap was divided into the lateral hinge and the four zones. b Sagittal multiplanar reconstruction computed tomographic (MPR-CT) images of each zone. The zone in which gap filling is observed on sagittal MPR-CT is the same as the zone in which trabecular bone continuity can be observed on anteroposterior radiography

Full size image

Risk factors for delayed bone healing

Delayed union was defined as no evidence of bone union in the radiographic evaluation at 3 months postoperatively. We investigated the following factors affecting bone formation: age, BMI, diabetes mellitus (DM), opening width, lateral hinge fracture, and direction of osteotomy line. Lateral hinge fracture was evaluated according to Takeuchi’s classification [24]: type I, the fracture reaches just proximal to or within the tibiofibular joint; type II, the fracture reaches the distal portion of the proximal tibiofibular joint (PTFJ); and type III, lateral plateau fracture. According to Han’s report [6], we also determined if the direction of the osteotomy line was within the safe zone (between the tip and circumference line of the fibular head) using postoperative AP (Fig. 3).

Fig. 3
figure 3

The safe zone (between the tip and circumference line of the fibular head)

Full size image

Statistical analysis

JMP version 11 (SAS Institute Inc., Cary, NC) was used to analyze and manage the data. The data are presented as means and standard deviations. Student’s t test was used to analyze quantitative data, and the chi-squared test was used to analyze qualitative data. Logistic regression analysis was performed to evaluate risk factors for delayed bone formation. A receiver operating characteristic (ROC) curve analysis was used to identify the opening width cut-off value for delayed bone healing. P < 0.05 was considered as indicative of statistical significance. The effect size 0.15, significant level of 0.05 and 80% power of the study on multiple regression analysis suggested a sample size of 99.

Results

Progression of osteotomy gap filling

Figure 4 shows the progression of osteotomy gap filling in 102 knees. The lateral hinge united at 3 months after surgery in 92 (90.2%) knees. At 1 year after surgery, 98 (96.1%) knees reached zone 1 and 92 (90.2%) knees reached zone 2. At plate removal, gap filling reached zone 2 in all cases and progressed further after plate removal. Loss of correction after plate removal was not observed in any case. The changes in knee alignment are shown in Table 2.

Fig. 4
figure 4

Progression of osteotomy gap filling in 102 knees. The vertical axis represents the ratio of cases reaching each zone. The lateral hinge united at 3 months after surgery in 92 (90.2%) knees. At 1 year after surgery, 98 (96.1%) knees reached zone 1 and 92 (90.2%) knees reached zone 2. At plate removal, gap filling reached zone 2 in all cases and progressed further after plate removal

Full size image
Table 2 Changes in knee alignment
Full size table

Risk factors for delayed bone healing

Risk factors for delayed bone healing are summarized in Table 3. In this study, there were 10 cases of delayed union. Two of the 10 cases underwent revision surgery because of nonunion. In addition, loss of correction occurred in 2 cases, and limitation of weight bearing until hinge union was needed in other 6 cases. No significant differences in age, BMI, DM, and lateral hinge fracture (Takeuchi’s classification type I) were observed between the patients with normal union and those with delayed union. Opening width (odds ratio: 1.61, P = 0.02), Takeuchi’s classification type II lateral hinge fracture (odds ratio: 20.4, P < 0.01), and osteotomy line below the safe zone (odds ratio: 8.98, P < 0.01) significantly delayed bone healing after OWHTO. ROC curve identified the opening width cut-off values for delayed bone healing to be 13.0 mm (AUC = 0.67, sensitivity = 60.0%, specificity = 66.7%).

Table 3 Risk factors for delayed bone healing
Full size table

Discussion

The most important findings of this study were as follows: (1) gap filling progressed from lateral to medial after OWHTO without bone grafting and progressed further after plate removal without loss of correction and (2) opening width over 13.0 mm, lateral hinge fracture (Takeuchi’s classification type II), and osteotomy line below the safe zone were risk factors for delayed bone healing.

The definition of gap filling is essential when evaluating the progression of bone formation in the osteotomy gap. Although MPR-CT is the most adequate method for assessing bone formation after medial OWHTO, it is impractical to use MPR-CT repeatedly. In the present study, we divided the osteotomy gap into the lateral hinge and the four zones on AP radiographs. Then, we compared bone formation in the osteotomy gap on AP radiographs and sagittal MPR-CT performed after plate removal. We found that the zone in which gap filling was observed on sagittal MPR-CT was the same as the zone in which trabecular bone continuity could be observed on AP radiographs. Therefore, we defined the zone in which trabecular bone continuity can be observed on AP radiographs as gap filling.

There have been few reports that have evaluated the progression of bone formation in the osteotomy gap after OWHTO without interposition material. Staubli et al. reported that standard radiography showed that ≥ 75% of the gap filled with new bone within 6 months. However, the CT at this point showed progression of bone mineralization, but no sign of full consolidation, which indicated that bone healing at 6 months was overestimated on standard radiography [23]. Brosset et al. divided the osteotomy gap into five zones to evaluate the progression of bone formation [1]. They reported that primary bone union occurred in 49 (96%) of 51 knees at an average of 4.5 months; 40 knees reached zone 4 at 4 months; and 43 (84%) knees reached zone 5 at 2 years. However, the gap filling on AP radiographs was not defined clearly, and measurement reliability tests were not performed in their reports. The inter- and intra-observer reliabilities of gap filling measurements in our study were high or almost perfect. Therefore, our evaluation method enabled us to assess the progression of gap filling reliably on AP radiographs without using MPR-CT.

The results of the present study demonstrated that bone formation progressed from the lateral hinge to the medial direction after OWHTO without filling the gap with bone graft or bone substitute and that gap filling progressed further after plate removal without loss of correction (Figs. 4, 5). The TomoFix plate was developed as an internal fixator to achieve optimal stability and maintain the desired correction without bone filler [10, 22]. Staubli reported that elasticity of the plate could also prove to be an important factor in encouraging bone growth inside the wedge by mechanical stimulation according to Wolff’s law [16, 23]. Röderer et al. reported that the use of dynamic locking screws in combination with early full load-bearing in OWHTO would increase interfragmentary movement and tissue strain and potentially induce good bone healing underneath the plate [17]. Shima et al. performed finite element analysis to evaluate stress and strain distribution under applied loading conditions and the relationship between mechanical stress and bone formation in the osteotomy gap after OWHTO with a TomoFix plate [20]. They demonstrated high strain energy at the lateral hinge after OWHTO and higher strain energy at the medial osteotomy gap after plate removal. The researchers also concluded that the high strain energy at the lateral hinge enhanced bone formation and gap filling and the higher strain energy at the medial osteotomy gap explained further gap filling after plate removal. These results agree with our clinical results. Therefore, we think that strain energy is an important factor in encouraging bone growth and that OWHTO using a TomoFix plate without bone grafting creates a favorable biomechanical environment to enhance bone formation in the osteotomy gap.

Fig. 5
figure 5

Radiographs of a 44-year-old man with osteoarthritis of the left knee. The lateral hinge united at 3 months after surgery. Thereafter, gap filling progressed from lateral to medial. After plate removal, gap filling progressed further

Full size image

Although the TomoFix plate provides high initial stability without bone grafting, its bulky design causes local discomfort in most patients. Plate removal is required in such cases, but there is a lack of information regarding the optimal timing for plate removal after OWHTO. Our results demonstrated that it is possible to remove the plate without loss of correction if gap filling reaches zone 2 on AP radiographs. We think that the mechanical stability of the osteotomy site is sufficient to prevent the loss of correction due to the bone union of the flange as well as gap filling to zone 2 even if the osteotomy gap is not completely filled.

In the present study, the occurrence of delayed- and nonunion were 8 (7.8%) and 2 knees (2.0%), respectively. These were almost equivalent to the respective rates of delayed- and nonunion of 6.6–15% and 1.6–7.0% observed in previous reports [1, 8, 11, 26]. In addition, opening width over 13.0 mm, lateral hinge fracture (Takeuchi’s classification type II), and osteotomy line below the safe zone were found to be risk factors for delayed bone formation. Regarding bone healing after fracture, various contributing factors leading to nonunion have been described, including smoking, drug or alcohol abuse, obesity, advanced age, poor bone quality, excessive motion at the fusion site, infection, reduced perfusion, metabolic disease (such as DM), and implant failure [5, 9]. Previous studies have identified obesity, smoking, the size of osteotomy gap, and fracture of the lateral hinge as risk factors that negatively influence bone healing after OWHTO, which potentially leads to delayed union or nonunion [11, 18, 21]. Larger osteotomy gap increases the potential of the lateral hinge fracture compared with smaller gap, which could lead in delayed bone healing. However, there is a lack of a strong evidence to establish clear guidelines in terms of the acceptable osteotomy size in OWHTO using TomoFix plate. Slevin et al. reported in a systematic review that OWHTO with gaps smaller than 10 mm and rigid fixation might be successfully managed without bone grafting [21]. Lobenhoffer et al. recommended filling of the osteotomy gap with autogenous cancellous bone in gaps over 13 mm in OWHTO using TomoFix plate [10]. Conversely, Schröter et al. reported that no correlation between the gap size and gap filling rate was found [18]. In their report, the mean gap size (9.4 ± 2.3 mm) was relatively smaller than the present study (12.0 ± 2.1 mm), which could lead in differing results. The present study demonstrated that the size of the osteotomy width over 13.0 mm could delay bone healing. Regarding the lateral hinge fracture, it has been reported that instability at the opening-wedge osteotomy site because of disruption of the lateral cortex can result in delayed union or nonunion and contribute to displacement and recurrent varus deformity [12, 13, 25]. Staubli et al. reported no progression of gap filling without union of the lateral hinge [22]. Fractures of the lateral hinge were divided into three types by Takeuchi et al. According to their reports, type II and type III lateral hinge fractures resulted in marked instability at the osteotomy site, whereas type I fractures were relatively stable because the soft tissue near the proximal tibiofibular joint (PTFJ) area is dense and solid [24]. Similarly, Schröter et al. reported that type II lateral hinge fractures delayed gap filling after OWHTO, whereas type I lateral hinge fractures did not affect bone healing, which is consistent with our results [18].

To prevent this unstable type II lateral hinge fracture in OWHTO, the direction of the osteotomy line is an important factor that must be considered. Han et al. defined the safe zone (between the tip and circumference line of the fibular head) for preventing unstable hinge fractures [6]. They described a high risk of type II fracture for osteotomy below the safe zone. Similarly, Nakamura et al. reported that the risk of unstable fractures was significantly lower in zone WL (within the PTFJ, lateral to the medial margin of the PTFJ) [14]. The osteotomy line below the safe zone was found to be a risk factor for delayed bone healing in the present study. Therefore, precise surgical technique, including appropriate direction of the osteotomy line and hinge position to maintain the integrity of the lateral cortex is the most important factor to prevent delayed union, loss of correction, and poor outcome.

There were several limitations in this study. First, bone formation on AP radiography and CT were compared only at plate removal. Although CT is the most suitable method for assessing bone healing after OWHTO, it is impractical to perform CT repeatedly. Second, there are concerns as to whether the radiographs tube was tilted into the plane of the osteotomy gap for precise evaluation. However, we confirmed neutral rotation and complete extension of the knee joint to obtain images suitable for evaluation. Third, we did not assess the bone union of the flange as well as the posterior cortex on lateral radiographs; however, it is difficult to evaluate the bone union of the flange on lateral radiographs because of the shade of the TomoFix plate. Despite these limitations, this study is, to our knowledge, the first to evaluate bone formation in the osteotomy gap after OWHTO, including after plate removal, on the basis of a clear definition of gap filling.

The clinical relevance of this study was that precise surgical techniques, including the appropriate direction of the osteotomy line and hinge position to maintain the integrity of the lateral cortex are the most important factor to prevent delayed union after OWHTO. In addition, gap filling progresses from lateral to medial after OWHTO without bone grafting. However, large opening gaps over 13.0 mm could delay bone healing. Therefore, in cases of gaps over 13.0 mm, autogenous bone grafting into osteotomy gaps is a considered a better option.

Conclusion

Gap filling progressed from lateral to medial after OWHTO without bone grafting and progressed further after plate removal. Large opening gaps, unstable hinge fractures, and osteotomy line below the safe zone cause delayed bone healing after OWHTO.