If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Anteromedial plating without filling the gap in open wedge high tibial osteotomy may increase the risk of screw breakage, which can be reduced by medial plating and bone-substitute insertion
Open wedge high tibial osteotomy (OWHTO) brings new complications such as screw breakages with or without correction loss and time-dependent increased posterior tibial slopes (PTS) due to the opening gap. For preventive purposes, we changed our OWHTO procedure from anteromedial plating without filling the gap (non-grafted group, n = 40, 2009–2012) to medial plating with bone-substitute insertion (grafted group, n = 45, 2012–2015). The objectives of this study were to evaluate the complication patterns and the effects of the modifications.
Methods
Patients undergoing OWHTO with TomoFix plates were included in this retrospective study. Demographics, clinical outcomes (flexion range and Japanese Orthopaedic Association score) and radiological outcomes (femorotibial angle) were assessed in both groups pre-operatively and 2-years postoperatively. The plate installation angle [PIA] and screw insertion depth [SID] were measured on computed tomographic slices at 6 months. PIA/SID was defined as the angle between the tibial anteroposterior axis and plate–width axis/the distance between the proximal screw tip and the proximal tibiofibular joint. The non-grafted group was further divided into complication and non-complication subgroups. Screw breakages were assessed during plate removal (1.5–2.5 years postoperatively).
Results
There were no differences in baseline characteristics or radiological/clinical outcomes between the non-grafted and grafted groups. There were 0 and 11 complications in the grafted and non-grafted groups, respectively. Complications included 7 screw breakages, 4 correction losses, and 5 time-dependent increased PTS with some overlaps. PIA and SID were significantly lower (p < 0.001) and higher (p < 0.001), respectively, in the grafted group and significantly lower (p = 0.018) and higher (p = 0.040), respectively, in the non-complication subgroup within the non-grafted group. The cutoff value of PIA for complications was calculated at 48.0°.
Conclusions
Medial plating OWHTO (PIA<48°) using bone-substitute with deeper screw insertion reinforces the opening gap for better angular stability compared with anteromedial plating without bone-substitute.
1. Introduction
Since the introduction of locking plates specifically designed for open wedge high tibial osteotomy (OWHTO), such as the TomoFix Medial High Tibial Plate (Synthes GmbH; Solothurn, Switzerland), the clinical outcomes of OWHTO have been significantly improved [
]. The improved clinical results may be attributed to the characteristic symmetry of the TomoFix, which provides greater flexibility (Fig. 1), and the pure titanium long plate. However, some new complications that did not exist for closed wedge high tibial osteotomy have arisen in OWHTO. The most common intra- and post-operative OWHTO-specific complications are lateral hinge fracture [
Large opening gaps, unstable hinge fractures, and osteotomy line below the safe zone cause delayed bone healing after open-wedge high tibial osteotomy.
Fig. 1The TomoFix plate is symmetric about its width. The flexibility of the plate is highest in the plane formed by the plate–length axis and the plate–thickness axis.
] because the plate is not routinely removed in European countries. In Asian individuals who underwent OWHTO with TomoFix, the plate was always removed [
] because it was too bulky to be left permanently in place. For this reason, we removed the plate in all OWHTO cases, which provided us with many more opportunities to inspect cases with screw breakages (Fig. 2) and/or increased posterior tibial slopes over time (Fig. 3). The increased posterior tibial slope over time is a lesser-known complication [
]. However, careful radiological evaluation after plate removal highlighted the importance of time-dependent slope change.
Fig. 2A 68-year-old female with multiple screw breakages complicated by a correction loss (Case 10). (A) Full-length view at 1 month reveals a mechanical medial proximal tibial angle (mMPTA) of 92°. Joint line (white line), mechanical axis (white dashed line). (B) Full-length anteroposterior view 2 years after surgery. A correction loss has occurred because of the change in the mMPTA (decreased to 88°). Screw breakage could not be radiologically detected. (C) The removed plate. Three out of four proximal screws (B, C, and D) were broken at the head–shaft interface.
Fig. 3A 69-year-old male with time-dependent increased posterior tibial slope (Case 1). The white circles indicate the posterior tibial slope. The slope was 3° at 1 month (A) and 8° at 2 years (B). Medial joint line (white line), the perpendicular line (white arrow) to the posterior cortex (white dashed line).
When we began using OWHTO with a locking plate, we did not insert any bone-substitutes (non-grafted group) according to our hypothesis that the stability of the locking system is enough for early full weight-bearing. Despite the controversy about the requirement for gap-fillers [
Opening wedge high tibial osteotomy performed without filling the defect but with locking plate fixation (TomoFix) and early weight-bearing: prospective evaluation of bone union, precision and maintenance of correction in 51 cases.
], we started to use bone substitutes (grafted group) to prevent complications after some experience of screw breakage and/or increased posterior tibial slope over time. Furthermore, some modifications of the plate installation (Fig. 4, Fig. 5, Fig. 6, Fig. 7) were made simultaneously with the commencement of bone-substitution insertion.
Fig. 4(A) Anteromedial plating generates the pendular micro-motion (grey arrow) of the fragment by the posterior shift of the contact point with flexion (grey and black triangles). A black bold arrow indicates the stress concentration at the head–shaft interface. (B) The pendular micro-motion of the proximal fragment is impeded (dashed arrows) by the bone-substitute wedges between the medial cortices. The more medially installed plate with screws normal to the tibial anteroposterior axis may experience less stress from pendular micro-motion.
Fig. 5Computed tomography measurement. The slice including three proximal screws (A) and the medial border of the proximal tibiofibular joint (B) are superimposed (C). The white triangle and the black double arrow show the plate installation angle and the screw insertion depth.
Fig. 6Axial cross-section of the knee. (A) Anteromedial plating in the non-grafted group. A bone spreader maintains the gap and the medial collateral ligament is reflected posteriorly. The screw depth indicates a negative value in most cases. (B) Medial plating in the grafted group. The anterior and posterior bone-substitute wedges maintain the gap. The medial collateral ligament is returned to the original position and repaired. The screw depth indicates a positive value in most cases.
Fig. 7Anteroposterior view. The dashed line, white line and the white arrow demonstrate the fibular contour, the medial margin of the proximal tibiofibular joint and the proximal screw-tips, respectively. (A) Anteromedial plating without bone-substitute and (B) medial plating with bone-substitute. The screw-tips are supported by the fibula in (B).
The first objective of this study was to compare the rate of delayed onset bone/implant-related complications such as screw breakage with and without correction loss and time-dependent increase of the posterior tibial slope between grafted and non-grafted groups. The second objective was to clarify the effect of the plate installation angle and the screw insertion depth on the rate of bone/implant-related complications.
2. Materials and methods
2.1 Cohorts
This retrospective cohort study was approved by the institutional review board of the authors’ affiliated institutions (no. 2020-2) and carried out in accordance with the Declaration of Helsinki. Written informed consent for this research was not obtained because of the retrospective design. The study population consisted of patients meeting the eligibility criteria [
] and who underwent OWHTO for medial compartment knee osteoarthritis. The cohorts were (1) OWHTOs treated between June 2009 and May 2012 without bone-substitute insertion (non-grafted group, i.e., pre-modification), and (2) OWHTOs treated between June 2012 and September 2015 with bone-substitute insertion (grafted group, i.e., post-modification).
2.2 Transition of indication and/or surgical procedure
During the study period, in addition to whether bone-substitute was used, there were several changes in our OWHTO indications and/or surgical procedure. In the early stages of the consecutive series, an opening distance <15 mm was considered to be an indication for OWHTO [
], an increased number of patients underwent a minor correction of <8 mm in the latter stage of this series. In terms of plate fixation, we performed OWHTOs using several types of locking plates: TomoFix (original, small, or new-generation standard versions) or TriS (Olympus Terumo Biomaterials, Tokyo, Japan) [
]. Although all plates were removed after a period of approximately 2 years, the number of removals occurring earlier than 1.5 years increased after the introduction of the bone-substitute.
2.3 Exclusion and inclusion criteria of this study
The patients who underwent OWHTO in combination with other surgeries (ligament reconstruction, osteochondral graft, or autologous chondrocyte implantation) and OWHTO with unstable lateral hinge fracture (Takeuchi type II or III) [
] were first excluded because these conditions have some potential adverse effect on the results. Furthermore, to compare both groups under the same conditions, only the cases that met the following inclusion criteria were included: 1) OWHTO was performed by a single surgeon, 2) small size TomoFix was used, 3) time to plate removal was between 1.5 and 2.5 years, and 4) opening distance was 8–12 mm. A flowchart illustrating the patient selection process is depicted in Fig. 8. The follow-up rate of the selected patients was 100%.
] with complete release of the superficial medial collateral ligament (MCL). The same rehabilitation was applied to both groups. Patients began their range of motion exercises within 2 days of their operation. Partial and full weight-bearing began 1 and 3 weeks after surgery, respectively.
2.5 Differences in surgical procedure between non-grafted and grafted groups
], a bone spreader was inserted into the posterior gap after the osteotomy, and the gap was opened to the distance determined by the target post-operative alignment. The MCL was released and reflected posteriorly, and the plate was installed on the anteromedial aspect (Fig. 4, Fig. 6, Fig. 7A) [
Bone formation and resorption in patients after implantation of beta-tricalcium phosphate blocks with 60% and 75% porosity in opening-wedge high tibial osteotomy.
J Biomed Mater Res B Appl Biomater.2008 Aug; 86: 453-459
]. Two bone-substitute wedges—posterior and anterior—were cut beforehand from the block to fit the target opening gap (Fig. 4, Fig. 6, Fig. 7B). After widening the posterior gap to the target distance, an additional bone spreader was inserted into the anterior gap to maintain the appropriate distance. The posterior spreader was then removed, and the posterior wedge was inserted. The anterior wedge was inserted after removing the anterior spreader. The MCL was returned into its original position beneath the pes anserinus and repaired (Fig. 6B). A TomoFix plate was installed on the medial aspect of the proximal tibia [
], with the proximal screws pointing toward the proximal tibiofibular joint (PTFJ) (Fig. 4, Fig. 5, Fig. 6, Fig. 7B).
2.6 Radiological examinations
Radiological evaluations comprised full-length anteroposterior views of the leg (pre- OWHTO, 1 month and 2 years post-OWHTO), anteroposterior/lateral views of the knee (pre- OWHTO, 1, 3, and 6 months, and 1, 1.5, and 2 years post-OWHTO), and computed tomography (CT) scans at 6 months post-OWHTO.
] were measured on the full-length radiograph and the posterior tibial slope was measured on the lateral view. The posterior tibial slope is defined as the angle between the medial joint line and the perpendicular line to the posterior cortex of the tibia because the view of the anterior cortex was impeded by the plate in some instances [
]. Correction loss was defined a reduction in the mechanical medial proximal tibial angle at 2 years compared with that at 1 month ≥ 3°. Time-dependent increased posterior tibial slope was defined as an increased slope at 2 years compared with that at 1 month ≥ 3°.
On the CT scans, using superimposed views of axial slices at 6 months, the plate installation angle and screw insertion depth were measured (Fig. 5, Fig. 6). The plate installation angle was defined as the angle between the tibial anteroposterior axis according to Akagi et al. [
] and the plate–width axis (Fig. 5, Fig. 6). The screw insertion depth, with reference to the PTFJ, was defined as the distance between the lateral tip of the screw and the margin of the PTFJ (Fig. 5, Fig. 6). When the screws were inserted beyond or short of the PTFJ, the values were defined as positive and negative, respectively (Fig. 5, Fig. 6).
2.7 Assessment
We collected characteristics of the cases including sex, age, body mass index, opening distance, and time from OWHTO to plate removal. Clinical outcomes were assessed by the Japanese Orthopaedic Association (JOA) score and the knee flexion range pre-OWHTO and 2 years post-OWHTO. Radiological data consisted of the femorotibial angle, the posterior tibial slope, the plate installation angle, and the screw insertion depth. Among the three types of bone/implant–related complications, correction loss and time-dependent increase in the posterior tibial slope were assessed at the designated radiological follow-up period mentioned above, and screw breakages were recorded at the plate-removing surgery independently of the radiological follow-up. The anterior, middle, and posterior screws of the plate were defined as A, B, and C, respectively, and the screw in the second row was defined as D (Fig. 2).
2.8 Statistical analysis
We compared demographic data, clinical data, radiological data, and complications between non-grafted and grafted groups. Chi-square tests and Fisher exact tests were applied to compare sex and the rate of complications. Student's t-tests, Mann–Whitney U tests, and Welch t-tests were used to compare the other values between groups. The non-grafted group was further divided into no bone/implant-related complications (non-complication group) and those with bone/implant-related complications (complication group) to reveal the solitary effect of the plate installation without the possible effect of the bone-substitute. The same evaluation items as in the non-grafted and grafted groups were compared between the non-complication and complication groups. Spearman's rank correlations between the plate installation angle and the screw insertion depth were calculated. Paired t-tests and Wilcoxon signed-rank tests were used to compare pre-operative and 2-year follow-up JOA scores, knee flexion ranges, and femorotibial angles. A p-value of <0.05 was considered to be statistically significant.
Logistic regression analysis was performed with the presence or absence of complications as the dependent variable, and the plate installation angle and screw insertion depth as the independent variables. A receiver-operating characteristic (ROC) curve and cutoff value were calculated after logistic regression analysis on intraoperative factors (plate installation angle and screw insertion depth). SPSS Statistics version 24 (IBM Corp., Armonk, New York) and R version 4.1.0 were used as the analysis software. A priori power analysis was conducted (G∗power, version 3.1.7) based on a medium effect size, and a post hoc power analysis was performed on the results of statistical analysis in each phase.
3. Results
Among 284 OWHTOs, 85 knees met the eligibility criteria (Fig. 8). The non-grafted and grafted groups comprised 40 and 45 cases, respectively (Table 1). Patients in the non-grafted group were not characteristically different from those in the grafted group; sex, age, body mass index, opening distance, and time to plate removal did not differ significantly between the groups (Table 1). Clinical and radiological parameters did not differ significantly between groups either pre-operatively or at 2 years (Table 2). In both groups, parameters were corrected by OWHTO, which is demonstrated by the significant differences between before OWHTO and 2 years after OWHTO (p < 0.001) (Table 2).
Table 1Comparisons between non-grafted and grafted group characteristics.
Characteristic
Non-grafted group n or mean (SD, range)
Grafted group n or mean (SD, range)
Between-group p value
Number of patients
40
45
N/A
Sex (male/female)
10/30
15/30
0.352
Age
62.2 (SD 8.1, 42 to 74)
65.3 (SD 8.4, 46 to 80)
0.082
Body mass index (kg/m2)
24.2 (SD 2.3, 20 to 30)
24.4 (SD 2.7, 20 to 32)
0.993
Opening distance (mm)
10.4 (SD 1.1, 8 to 12)
9.9 (SD 1.3, 8 to 12)
0.080
Time to plate removal (year)
2.0 (SD 0.1, 1.6 to 2.2)
1.9 (SD 0.2, 1.6 to 2.4)
0.092
The chi-square test was used to compare sex; the Student's t-test was used to compare age, and the Mann–Whitney U-test was used to compare body mass index, opening distance, and time to plate removal between the non-grafted and grafted groups.
Table 2Comparisons between non-grafted and grafted group outcomes and assessments.
Characteristic
Assessment
Non-grafted group mean (SD, range)
Comparison between assessment p value
Grafted group mean (SD, range)
Comparison between assessment p value
Between-group p value
Japanese Orthopaedic Association score
Pre-operative
69.5 (SD 12.2, 40 to 85)
<0.001
69.2 (SD 10.6, 40 to 85)
<0.001
0.785
Two years
91.8 (SD 8.4, 60 to 100)
93.6 (SD 5.5, 80 to 100)
0.501
Knee flexion range (°)
Pre-operative
133.3 (SD 13.9, 100 to 155)
<0.001
138.0 (SD 10.5, 120 to 155)
<0.001
0.164
2 years
146.0 (SD 8.8, 125 to 155)
145.0 (SD 8.9, 130 to 155)
0.579
Femorotibial angle (°)
Pre-operative
179.2 (SD 1.8, 174 to 182)
<0.001
178.5 (SD 2.5, 172 to 184)
<0.001
0.139
2 years
170.9 (SD 2.0, 163 to 174)
171.1 (SD 3.0, 166 to 182)
0.585
Plate installation angle (°)
6 months
47.9 (SD 7.2, 25 to 63)
–
26.2 (SD 10.6, 4 to 47)
–
<0.001
Screw insertion depth (mm)
6 months
−5.0 (SD 4.9, −17 to 7)
–
4.3 (SD 4.2, −7 to 11)
–
<0.001
Bone/implant-related complications
1.5–2.5 years (at the time of plate removal)
11
–
0
–
<0.001
Paired t-tests and Wilcoxon signed-rank tests were used to compare pre-operative and 2-year follow-up values. Mann–Whitney U-tests and Welch t-tests were used to compare Japanese Orthopaedic Association score, knee flexion range, femorotibial angle, plate installation angle, and screw insertion depth between the non-grafted and grafted groups. The chi-square test was used to compare the rates of bone/implant-related complications.
The plate installation angle was significantly lower (p < 0.001) and the screw insertion depth was significantly higher (p < 0.001) in the grafted group (Table 2). When the plate installation angle and the screw insertion depth were plotted, there was a strong linear relationship between them (R2 = 0.537, p < 0.001) (Fig. 9).
Fig. 9Relationship between screw insertion depth and the plate installation angle. The cutoff value of the plate installation angle for the risk of bone/implant-related complications is denoted by the dashed line.
There were no complications in the grafted group, and the complication rate was significantly higher in the non-grafted group than that in the grafted group (p < 0.001) (Table 2). Post hoc power analysis for this outcome (alpha value of 0.05, Cohen's d of 2.04 in screw insertion depth, and total sample size of 85 for a 2-tailed hypothesis) demonstrated an observed power of 100%.
In the non-grafted group, there were 11 cases with bone/implant-related complications (Table 3) and Case 9 was converted to total knee arthroplasty. The complications consisted of 7 screw breakages, 4 correction losses, and 5 time-dependent posterior tibial slope increases with some overlaps (Table 3). All screw breakages occurred at the head/shaft interface (Fig. 2C). Thirteen screws were broken, with 0, 4, 6, and 3 breakages occurring in A-, B-, C-, and D-screws, respectively (Table 3). Three out of four correction loss cases had multiple broken screws, while four out of five cases with increased posterior tibial slopes had no broken screws (Table 3). When the non-grafted group was divided into non-complication and complication groups, demographics, pre-operative clinical measures, and pre-operative radiological measures did not differ significantly between groups (Table 4, Table 5); however, the femorotibial angle at 2 years was significantly larger for the complication group because there were some correction losses (Table 4, Table 5). Nevertheless, the JOA score and knee flexion range at 2 years did not differ significantly between the non-complication and complication groups (Table 4, Table 5).
Table 3Complications.
Case
Screw breakage
Mechanical medial proximal tibial angle (°)
Correction loss
Posterior tibial slope (°)
Time-dependent increased posterior tibial slope
1 month
2 years
1 month
2 years
1
0
91
91
–
3
8
+
2
0
90
89
–
8
13
+
3
0
92
92
–
6
10
+
4
0
90
89
–
6
9
+
5
C
90
89
–
7
7
–
6
C
97
97
–
4
6
–
7
B,C
91
89
–
8
9
–
8
B,C
92
88
+
6
11
+
9
B,C,D
95
88
+
7
8
–
10
B,C,D
92
88
+
7
7
–
11
D
94
89
+
0
0
–
0, no breakage, or breakage of plate screws. A, anterior screw; B, middle screw; C, posterior screw; D, lower middle screw.
+indicates a reduction in mechanical medial proximal tibial angle ≥3°or an increase in posterior tibial slope ≥3°.
- indicates a reduction in mechanical medial proximal tibial angle <3°or an increase in posterior tibial slope <3°.
Table 4Comparisons between characteristics of the non-complication and complication subgroups in the non-grafted group.
Characteristic
Non-complication group n or mean (SD, range)
Complication group n or mean (SD, range)
Between-group p value
Number of patients
29
11
N/A
Sex (male/female)
8/21
2/9
0.432
Age
62.7 (SD 8.0, 42 to 74)
60.9 (SD 8.7, 45 to 73)
0.538
Body mass index (kg/m2)
23.7 (SD 2.2, 20 to 30)
25.5 (SD 2.1, 22 to 30)
0.077
Opening distance (mm)
10.3 (SD 1.2, 8 to 12)
10.5 (SD 0.7, 9 to 11.5)
0.785
Time to plate removal (years)
2.0 (SD 0.1, 1.6 to 2.1)
2.0 (SD 0.1, 1.7 to 2.2)
0.203
The Fisher exact test was used to compare sex, Student's t-test was used to compare age and body mass index, and the Mann–Whitney U-test was used to compare opening distance and time to plate removal between the non-complication and complication groups.
Table 5Comparisons between outcomes and assessments of the non-complication and complication subgroups in the non-grafted group.
Characteristic
Assessment
Non-complication group mean (SD, range)
Comparison between assessment p value
Complication group mean (SD, range)
Comparison between assessment p value
Between-group p value
Japanese Orthopaedic Association score
Pre-operative
68.3 (SD 11.5, 40 to 85)
<0.001
72.7 (SD 13.8, 50 to 85)
0.003
0.226
2 years
91.7 (SD 7.0, 80 to 100)
91.8 (SD 11.9, 60 to 100)
0.975
Knee flexion range (°)
Pre-operative
132.4 (SD 14.4, 100 to 150)
<0.001
135.5 (SD 12.7, 120 to 155)
0.003
0.543
2 years
146.1 (SD 8.1, 125 to 155)
145.9 (SD 10.9, 125 to 155)
0.960
Femorotibial angle (°)
Pre-operative
178.9 (SD 1.7, 174 to 182)
<0.001
180.2 (SD 1.6, 178 to 184)
0.005
0.056
2 years
170.2 (SD 2.2, 166 to 176)
173.5 (SD 3.6, 171 to 182)
0.002
Plate installation angle (°)
6 months
46.3 (SD 5.1, 25 to 58)
–
52.0 (SD 5.0, 45 to 63)
–
0.018
Screw insertion depth (mm)
6 months
−4.1 (SD 5.1, −17 to 7)
–
−7.4 (SD 3.5, −15 to −3)
–
0.040
Paired t-tests and Wilcoxon signed-rank tests were used to compare pre-operative and 2-year follow-up values. Mann–Whitney U-tests and Student's t-tests were used to compare Japanese Orthopaedic Association score, knee flexion range, femorotibial angle, plate installation angle, and screw length on the proximal tibiofibular joint between the non-complication and complication groups.
The plate installation angle was significantly lower (p = 0.018) and the screw insertion depth was significantly higher (p = 0.040) in the non-complication group than in the complication group (Table 4, Table 5). Logistic regression analysis identified the plate installation angle as a possible factor (standardized regression coefficient β = 0.302, odds: 1.18). In the analysis of the ROC curve, the cutoff value of the plate installation angle was 48.0, area under curve was 0.699, sensitivity was 0.817, and specificity was 0.621 (Fig. 10). Post hoc power analysis between these two groups (alpha value of 0.05, Cohen's d of 1.13 in the plate installation angle, and total sample size of 40) demonstrated an observed power of 86%.
Fig. 10Receiver-operating characteristic curve for the plate installation angle. The cutoff value is 48.0° and the area under curve is 0.699.
Because there were no systematic differences between the baseline characteristics of the non-grafted group and those of the grafted group (Table 1, Table 2), we consider it reasonable to compare outcomes between the groups. The results of comparison under the same conditions revealed numerous bone/implant–related complications that occurred only in the non-grafted group. Therefore, our initial hypothesis that the stability of the TomoFix is enough for early full weight-bearing without filling the gap was rejected.
However, simultaneous modification of the plate installation could be a possible complication factor that should be considered. The significantly smaller plate installation angle in the grafted group (Table 1, Table 2); that is, screw insertion normal to the tibial anteroposterior axis (Fig. 4, Fig. 5, Fig. 6), indicated that, with the modification, the plate was installed on the more medial aspect rather than the anteromedial aspect of the tibia. In addition to the intentional medial plating, plate fixation after removing the spreader made medial plating technically easier (Fig. 6). The significantly larger screw insertion depth in the grafted group (Table 1, Table 2) indicated that, with the modification, the proximal screws were aimed toward the PTFJ. The strong linear relationship between the plate installation angle and the screw insertion depth (Fig. 9) meant that more medial plating enabled deeper screw insertion towards the PTFJ. No bone/implant-related complications were found in the grafted group (Table 1, Table 2), which suggests that the modified surgical procedure may reduce the chance of complications.
In the non-grafted group, of the three bone/implant-related complications, the most frequent was screw breakage (Table 3). No screw breakage was caused by trauma; the breakages appeared to be caused by stress beyond the long-term durability of the screws. Interestingly, all cases except one included the posterior screw (Table 3), which suggests that the posterior part of the plate might be exposed to more stress than the other parts of the plate.
] also noted that the femoro-tibial contact point moves posteriorly. Furthermore, TomoFix has maximum flexibility because of its symmetry in the plane formed by the plate–length axis and the plate–thickness axis (Fig. 1) and the high elastic modulus of the pure titanium. Accordingly, in the non-grafted group, the posterior-dominant stress during walking and the anteriorly installed flexible plate may generate repetitive posterior-dominant pendular micro-motion of the proximal fragment (Fig. 1, Fig. 4 and Fig. 1, Fig. 4A). In other words, in cases with a larger plate installation angle (Fig. 6A), the locking screws would be more directly involved in the motion and the angular stability may decrease further (Fig. 4A). The involvement in the pendular micro-motion could induce screw breakage at the head/shaft interface (i.e., the leverage point of the motion). According to the cutoff value of the ROC curve (Fig. 10), a plate installation angle of <48° can be recommended to reduce the risk of delayed bone/implant-related complications. Furthermore, because no complications occurred in cases with a positive screw insertion depth, insertion of the screw tip above the PTFJ, which allows the fibula to work as a supporting strut [
] that the posterior stability of the osteotomy site might be weakened in the anteromedial position when TomoFix is used without a bone substitute. Thus, if long-term pendular micro-motion causes stress that surpasses screw durability, the posterior screw, with the largest arc, may experience breakage, while the anterior screw would be less affected (Fig. 4A). Breakage of the posterior screw may subsequently expose the middle and anterior screws to overload and result in further screw breakage.
Screw breakages seemed to occur prior to correction loss in the coronal plane; in two out of three cases without correction loss, only the posterior screw experienced breakage, while multiple screw breakages occurred in three out of four cases with correction loss (Table 3). When a single screw experiences breakage, the remaining three proximal screws may manage to support the medial side; however, with multiple screw breakages, the loss of medial support for the axial load is followed by correction loss.
The increased posterior tibial slope without screw breakage can be explained by the motion of lumber spine. Mizuno et al. [
] reported that repetitive movement in the lumbar spine might cause loosening of the screw after pedicle screw fixation, which can be observed as a clear zone formation around the screws [
]. When the cancellous bone in the proximal part of the osteotomy succumbs to the effects of pendular micro-motion of the screw, increased posterior tibial slope instead of screw breakage may occur (Fig. 11).
Fig. 11Clear zone formation around the proximal screws. The white lines indicate the sclerotic margin of the removed screws. Expanded proximal screw holes as a trace of the clear zone can be observed.
In the grafted group, even though OWHTO generates medial cortex discontinuity from the opening gap, a strut is created that comprises the proximal cortex, bone-substitutes, and the distal cortex (Fig. 4, Fig. 7 and Fig. 4, Fig. 7B). Because the OSferion 60, with 60% porosity, used here can provide adequate strength to weight-bearing [
Bone formation and resorption in patients after implantation of beta-tricalcium phosphate blocks with 60% and 75% porosity in opening-wedge high tibial osteotomy.
J Biomed Mater Res B Appl Biomater.2008 Aug; 86: 453-459
], the strut would reinforce the gap by reducing posterior sinking from the pendular micro-motion (Fig. 4). This continuity may redistribute weight-bearing to the strut from the TomoFix and improve the TomoFix's durability (Fig. 4, Fig. 7 and Fig. 4, Fig. 7B), thus helping to reduce complications. In addition to the effect of the bone-substitute itself, facilitating the medial plating towards above-PTFJ insertion (Fig. 7B) by modification may also assist in reducing complications.
Following the OWHTO modification, post-operative rehabilitation and lifestyle guidance was revised to minimize pendular micro-motion and is being investigated as an ongoing study. Walking exercises without bending the knee begin early after surgery. Passive flexion exercises and deep knee flexion, such as that experienced during Japanese-style sitting or weeding a garden in a squatting position, are allowed after bone union.
This study had some limitations. First, this was a retrospective study with a small sample size. Second, several modifications that could not be separated from each other were simultaneously performed. Therefore, it is difficult to understand how each modification—bone-substitute insertion, changes in the plate installation angle, or the screw insertion depth—contributed to the prevention of bone/implant–related complications.
In conclusion, anteromedial plating with a plate installation angle of >48° without filling the gap may increase the risk of bone/implant-related complications such as screw breakage, correction loss, and time-dependent increased posterior tibial slope. The complication risk can be reduced by medial plating with bone-substitute insertion.
IRB approval
This study was approved by our institutional review board (No.2020-2).
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Authors’ contributions
Each author made substantial contributions to the following components of the project: conception, design, interpretation of data, drafting of the manuscript. All authors read and approved the final manuscript.
Declaration of competing interest
Dr. Ryuichi Nakamura has a consultancy with Olympus Terumo Biomaterials. All other authors have no conflicts of interest relevant to this article to declare.
Acknowledgments
We thank Etsuko Tabata and Ippei Yoshida for data collection and statistical support with the manuscript, respectively.
References
Staubli A.E.
De Simoni C.
Babst R.
Lobenhoffer P.
TomoFix: a new LCP-concept for open wedge osteotomy of the medial proximal tibia--early results in 92 cases.
Large opening gaps, unstable hinge fractures, and osteotomy line below the safe zone cause delayed bone healing after open-wedge high tibial osteotomy.
Opening wedge high tibial osteotomy performed without filling the defect but with locking plate fixation (TomoFix) and early weight-bearing: prospective evaluation of bone union, precision and maintenance of correction in 51 cases.
Bone formation and resorption in patients after implantation of beta-tricalcium phosphate blocks with 60% and 75% porosity in opening-wedge high tibial osteotomy.
J Biomed Mater Res B Appl Biomater.2008 Aug; 86: 453-459
00006-9&id=gr1.jpg){kind=link}
00006-9&id=gr2.jpg){kind=link}
00006-9&id=gr3.jpg){kind=link}
00006-9&id=gr4.jpg){kind=link}
00006-9&id=gr5.jpg){kind=link}
00006-9&id=gr6.jpg){kind=link}
00006-9&id=gr7.jpg){kind=link}
00006-9&id=gr8.jpg){kind=link}
00006-9&id=gr9.jpg){kind=link}
00006-9&id=gr10.jpg){kind=link}
00006-9&id=gr11.jpg){kind=link}