Comparison of three fixation methods for posterolateral fragments in tibial plateau fractures: a biomechanical study

Abstract

Background

Tibial plateau fractures involving posterolateral fragments are challenging to fix internally because of the small surface area and the complex regional anatomy. This study aimed to compare the biomechanical performance of three plate fixation methods and to identify the construct most resistant to fracture displacement under axial loading.

Methods

Fifteen synthetic tibia models were used to simulate posterolateral split fractures. The models were randomized into three groups (n = 5 per group): Group A received an R-LOCK locking plate with 3.5 mm locking screws; Group B received a small T-oblique dynamic compression plate (DCP) with 3.5-mm cortical screws; and Group C received a distal radius T-oblique locking compression plate (LCP) with 2.7 mm locking screws. Vertical displacement was measured under sequential axial loads of 500, 1000, and 1500 N, followed by load-to-failure testing. Because each specimen was tested under multiple loading conditions, vertical displacement was analysed by mixed-design repeated-measures ANOVA (within-subject factor, loading condition; between-subject factor, implant group), equivalent in this balanced design to a random-intercept linear mixed-effects model with specimen as the random grouping factor. Per-load one-way ANOVAs were used as a secondary analysis for pairwise comparisons.

Results

For vertical displacement, the primary mixed-design repeated-measures ANOVA revealed a significant main effect of load (p < 0.001), only a marginal main effect of implant group (p = 0.077), and a non-significant group × load interaction (p = 0.289). Based on secondary per-load analyses, Group C demonstrated significantly less vertical displacement than Group A and Group B under low-to-moderate axial loading at both 500 N (0.219 ± 0.036 mm vs. 0.359 ± 0.092 mm and 0.329 ± 0.055 mm; p = 0.005 and p = 0.021, respectively) and at 1000 N (0.382 ± 0.053 mm vs. 0.494 ± 0.114 mm and 0.576 ± 0.089 mm; p = 0.005). At 1500 N, Group C continued to exhibit the lowest mean displacement (0.592 ± 0.165 mm) compared with Group A (0.601 ± 0.182 mm) and Group B (0.757 ± 0.218 mm), although this difference did not reach statistical significance (p = 0.336). No statistically significant difference in load-to-failure was observed among the three constructs (p = 0.875), although a descriptive trend favoured Group C (3260.40 ± 351.76 N), followed by Group B (3178.80 ± 218.88 N) and Group A (3145.80 ± 461.66 N).

Conclusions

We demonstrated that the 2.7 mm distal radius T-oblique LCP was associated with comparatively lower vertical displacement under moderate axial loads (500–1000 N) compared to Groups A and B. All three constructs maintained vertical displacement below the 2 mm failure threshold across the tested loading range, suggesting that each is biomechanically adequate to resist normal physiological loads.

Trial registration

This study used only synthetic tibial bone models and did not involve human participants, human tissue, or animals; therefore, formal ethical approval and informed consent were not required. The protocol was nevertheless reviewed and registered by the Institutional Review Board of the Faculty of Medicine Vajira Hospital, Navamindradhiraj University (COE 031/2022).

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12891-026-10140-4.

Background

Tibial plateau fractures are intra-articular fractures that require meticulous management to restore the articular surface, maintain lower-limb alignment, and allow for early range of motion. Current management strategies typically involve open reduction and internal fixation (ORIF) with plates and screws [1].

Attention to the management of posterolateral fragments in tibial plateau fractures has increased in recent years. Biomechanical studies have demonstrated that posterolateral buttress plate is an effective method of resisting fracture displacement [1–5]. However, because the surgical exposure of the posterolateral column is limited, implant placement is challenging, and no standard implant is currently designed specifically for this purpose [6, 7]. To date, different implant types have not been extensively compared with respect to their biomechanical performance in stabilizing these fractures.

By comparing the biomechanical performance of various implants that conform to the contour of the posterolateral surface of the tibial plateau, we sought to identify which fixation type provides the greatest stability against further fracture displacement. This information may help guide the selection of suitable implants for treating these fractures.

Methods

Study design

Fifteen synthetic right tibia models (Synbone type 1110; Synbone AG, Zizers, Switzerland) were used to simulate posterolateral tibial plateau fractures. These models are widely used in biomechanical research because their mechanical properties are comparable to those of human cancellous and cortical bone [8]. The specimens were randomly assigned to three equal fixation groups (n = 5 per group) using a web-based random sequence generator. The groups were defined as follows: Group A was stabilized with an R-LOCK locking plate using 3.5 mm locking screws. Group B was stabilized with a small T-oblique dynamic compression plate (DCP) using 3.5 mm cortical screws and Group C was stabilized with a distal radius T-oblique locking compression plate (LCP) using 2.7 mm locking screws. The objective of the study was to evaluate and compare the biomechanical performance of these three constructs under progressive axial loading and load-to-failure conditions. Because this in vitro biomechanical study used only synthetic bone models and did not involve human participants, human tissue, or animals, formal ethical review was not required; however, the study protocol was reviewed and registered by the Institutional Review Board of the Faculty of Medicine Vajira Hospital, Navamindradhiraj University (COE 031/2022)(Date of acknowledgement 30/09/2022).

Surgical simulation and fixation methods

A single surgeon (T.T.) created all fractures using a 10 mm oscillating saw blade. All models were subjected to posterolateral split osteotomies. The fracture morphology was based on the configuration described by Sohn et al. [9] and was defined by a sagittal angle (SA) of 75°. The osteotomies were performed such that the diagonal distance (DD) between the lateral and posterior cortical exit points and the average posterior cortical distance (PCD), both measured 30 mm. This configuration was maintained across all specimens to ensure uniformity (Fig. 1).

Fig. 1.

Creation of the posterolateral column fracture. The sagittal angle (SA) was 75°. The diagonal distance between the lateral and posterior exit points (DD) was 30 mm, and the average posterior cortical distance (PCD) was 30 mm (a, superior view; b, posterior view; c, lateral view)

All surgical fixations were performed by a single trauma-trained orthopaedic surgeon (K.L.). Although the fixation method varied by plate design, a consistent placement protocol relative to anatomical landmarks was strictly maintained within each group to minimize experimental variability. Group A: The posterolateral fragment was stabilized using an R-LOCK locking plate (Intercus, Germany) with 3.5 mm locking screws (Fig. 2a). Four screws were used: two locking screws(length, 45 mm) were directed through the posterolateral fragment into the anterolateral cortex for stability, and two screws (length, 35 mm) were placed in the metaphysodiaphyseal region. All locking screws were inserted perpendicularly via dedicated drill guides, and final tightening was performed using a calibrated torque-limiting driver to standardize fixation stability. Group B: The posterolateral fragment was stabilized using a small T-oblique dynamic compression plate (DCP) with 3.5 mm cortical screws (Fig. 2b). Five screws were inserted in total: three screws (length, 45 mm) were directed from the posterolateral fragment toward the anterolateral cortex, and two screws (length, 35 mm) were positioned in the metaphyseal area. Screws were inserted perpendicularly across the fracture site to achieve optimal compression and stability. Group C: The posterolateral fragment was stabilized using a distal radius T-oblique locking compression plate (LCP) with 2.7 mm locking screws (Fig. 2c). Five screws were used: three locking screws (length, 45 mm) were directed through the posterolateral fragment into the anterolateral cortex, and two screws (length, 35 mm) were placed in the metaphysodiaphyseal region (one compression screw and one locking screw). As in Group A, locking screws were inserted perpendicularly through drill guides and tightened with a calibrated torque-limiting driver to ensure consistent construct stability.

Fig. 2.

Three implant fixation types: a Tibial plateau fracture; posterolateral column; intra-articular fracture; biomechanical study. Group A was fixed with the R-LOCK locking plate (Intercus, Germany) with 3.5 mm locking screws; b Group B was fixed with the T-oblique dynamic compression plate (DCP) with 3.5 mm cortical screws; c Group C was fixed with the distal radius T-oblique locking compression plate with a 2.7 mm locking screw diameter

Biomechanical testing

Mechanical testing was performed using an Instron 5566 universal materials-testing machine (Instron, Norwood, MA, USA). The tibial models were osteotomized 15 cm distal to the proximal articular surface, and the distal segments were oriented vertically and embedded in denture base resin to ensure rigid fixation. All experiments were performed under controlled laboratory conditions at a constant ambient temperature of 25 °C.

Vertical displacement was quantified using two-point measurement along the vertical axis, captured by a high-definition, motion-sensing camera (GeniCam; IDS Imaging Development Systems GmbH, Obersulm, Germany). Image data were processed using ImageJ software (National Institutes of Health, Bethesda, MD, USA), with measurements recorded in millimetres to three decimal places.

Biomechanical evaluation was performed using a synthetic posterior femoral condyle to replicate physiological loading at 90° of knee flexion [10]. A compressive force was applied at a loading rate of 10 N/s. Axial peak loads of 500, 1000, and 1500 N were tested sequentially on each specimen (Fig. 3). The vertical displacement of the posterolateral fragment was recorded at each loading increment using the motion-sensing system described above. After the 1500 N measurement, the load was increased continuously until fixation failure occurred. In accordance with established literature, failure was defined as a vertical displacement of 2 mm [11]. The ultimate load required to reach this 2 mm displacement threshold was recorded as the load-to-failure.

Fig. 3.

Each specimen was positioned vertically on an Instron 5566 materials testing machine (Instron, Norwood, MA, USA). Axial loads of 500, 1000, and 1500 N were applied through a synthetic posterior femoral condyle at 90° of knee flexion to simulate physiological loading on the posterolateral tibial plateau

Statistical analysis

All data are presented as the mean ± standard deviation (mean ± SD). Because each specimen was tested under multiple sequential loading conditions (500, 1000, and 1500 N), displacement measurements were not fully independent. To account for this within-specimen correlation, a mixed-design repeated-measures ANOVA with the loading condition as the within-subject factor and the implant group as the between-subject factor was used as the primary analysis for vertical displacement; in this balanced design (n = 5 per group, no missing observations) this model is algebraically equivalent to a random-intercept linear mixed-effects model with specimen as the random grouping factor. To facilitate interpretation and pairwise comparisons at each load level, separate one-way ANOVAs were also performed at 500, 1000, and 1500 N, and for the load-to-failure outcome. When a one-way ANOVA revealed a significant main effect, Fisher’s Least Significant Difference (LSD) post hoc test was used for pairwise comparisons. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using Stata version 18 (StataCorp, College Station, TX, USA).

An a priori power analysis based on prior data [5] (f ≈ 0.7, α = 0.05, power = 0.80) required five specimens per group to detect differences at 500 and 1000 N; post hoc power was adequate at these loads but limited at 1500 N and for load-to-failure (Type II error cannot be excluded). Small sample size, the normality assumption, and residual within-specimen correlation from sequential loading are acknowledged as analytical limitations below.

Results

Vertical displacement

Vertical displacement under three axial loads is summarized in Table 1. Statistically significant differences in vertical displacement were observed at 500 N (p = 0.013) and at 1000 N (p = 0.016). At 500 N, Group C (0.219 ± 0.036 mm) exhibited significantly lower displacement than Group A (0.359 ± 0.092 mm) (p = 0.005) and Group B (0.329 ± 0.055 mm) (p = 0.021). At 1000 N, Group C (0.382 ± 0.053 mm) continued to demonstrate significantly lower displacement than Group A (0.494 ± 0.114 mm) and Group B (0.576 ± 0.089 mm) (p = 0.005). At 1500 N, Group C (0.592 ± 0.165 mm) showed lower mean displacement than Group A (0.601 ± 0.182 mm) and Group B (0.757 ± 0.218 mm), but this difference was not statistically significant (p = 0.336). The mixed-design repeated-measures ANOVA showed a significant main effect of loading condition (F(2, 24) = 46.97, p < 0.001), a marginal main effect of implant group (F(2, 12) = 3.20, p = 0.077), and a non-significant group × load interaction (F(4, 24) = 1.33, p = 0.289); per-load one-way ANOVAs are therefore reported as secondary descriptive analyses.

Table 1.

Vertical displacement of the posterolateral fragment under various axial loads

Implant Group	500 N Displacement (mm) (Mean ± SD)	1000 N Displacement (mm) (Mean ± SD)	1500 N Displacement (mm) (Mean ± SD)
Group A	0.359 ± 0.092	0.494 ± 0.114	0.601 ± 0.182
Group B	0.329 ± 0.055	0.576 ± 0.089	0.757 ± 0.218
Group C	0.219 ± 0.036	0.382 ± 0.053	0.592 ± 0.165
p-value (per-load, secondary)	0.013	0.016	0.336
Primary analysis (mixed-design RM-ANOVA)	Group: F(2, 12) = 3.20, p = 0.077; Load: F(2, 24) = 46.97, p < 0.001; Group × Load: F(4, 24) = 1.33, p = 0.289

Data are presented as mean ± standard deviation. Differences among groups at each load level were tested using one-way ANOVA. A mixed-design repeated-measures ANOVA—with load as the within-subject factor and implant group as the between-subject factor, equivalent to a random-intercept linear mixed-effects model—served as the primary analysis to account for within-specimen correlation

A p-value < 0.05 was considered statistically significant

Load-to -failure

As shown in Table 2, no statistically significant difference in ultimate load-to-failure was observed among the three constructs (p = 0.875). Despite the absence of statistical significance, a descriptive trend was observed in which Group C withstood the highest mean load-to-failure (3260.40 ± 351.76 N), followed by Group B (3178.80 ± 218.88 N) and Group A (3145.80 ± 461.66 N).

Table 2.

Load-to-failure for the three internal fixation methods

Parameter	Group A (n = 5)	Group B (n = 5)	Group C (n = 5)	p-value
Load-to-failure (N) (Mean ± SD)	3145.80 ± 461.66	3178.80 ± 218.88	3260.40 ± 351.76	0.875

Data are presented as mean ± standard deviation. Differences among groups were tested using one-way ANOVA

A p-value < 0.05 was considered statistically significant

Discussion

The management of posterolateral tibial plateau fractures remains surgically challenging owing to the limited operative exposure imposed by the surrounding neurovascular anatomy and the irregular contour of the posterior proximal tibial surface. Although posterior buttress plating is widely accepted as a biomechanically favourable fixation strategy for these injuries [1–4], there is currently no standard implant specifically designed for this anatomical region. The present study compared the biomechanical performance of three plate configurations—the 3.5 mm R-LOCK locking plate (Group A), the 3.5 mm small T-oblique dynamic compression plate (Group B), and the 2.7 mm distal radius T-oblique locking compression plate (Group C)—under progressive axial loading using standardized synthetic tibial plateau fracture models.

Our findings suggest that Group C was associated with lower vertical displacement than both Group A and Group B at 500 N, and with lower displacement than Group B at 1000 N. However, no statistically significant differences in displacement were observed among the three constructs at 1500 N, and no significant difference in load-to-failure was detected. Importantly, all three implants maintained vertical displacement below the 2 mm clinical failure threshold across all loading conditions up to 1500 N, indicating that each construct provided adequate mechanical stability within the tested physiological loading range.

The data suggest a load-dependent pattern. The smaller 2.7 mm implant appeared to offer a structural advantage at lower loads, but this advantage diminished as the applied force increased. The comparatively lower displacement of the 2.7 mm distal radius T-oblique LCP at low-to-moderate loads may be explained by its T-oblique geometry, which provides a closer contour match to the curved posterolateral surface of the proximal tibia than the flatter profiles of the other tested implants. This improved plate-to-bone congruence likely facilitates more uniform load distribution across the implant–bone interface, reducing localized stress concentration and limiting early micromotion at the fracture site. Improved congruency may also enhance the buttressing effect against shear forces. However, this mechanical advantage appeared load-dependent and was not sustained at higher loads. As axial load increased to 1500 N, intra-group displacement variability increased and inter-group differences were no longer statistically significant. This convergence at higher loads is most likely attributable to the inherently lower load-bearing capacity of the 2.7 mm locking screws used in Group C compared with the 3.5 mm screws used in Groups A and B. Under higher loads, the contouring advantage of Group C may have been offset by the reduced cross-sectional area of the smaller screws, yielding displacement comparable to that of the larger-screw constructs. These findings should therefore be interpreted as evidence of a relative advantage in the low-to-moderate loading range rather than as overall biomechanical superiority.

These results are broadly consistent with previous biomechanical investigations. Zhang et al. [5] reported that posterior buttressing plates provided stronger resistance to shear forces in posterolateral split fractures than lateral locking plates, a principle supported in the present study, in which all three posterior buttress constructs performed adequately. Hu et al. [1] reported that a horizontal belt plate conferred no biomechanical advantage over a 3.5 mm lateral LCP, reinforcing the view that plate position relative to the primary fracture displacement vector may be more important than plate size alone. Zhang et al. [5] also reported that a 2.7 mm distal radius locking plate produced less vertical displacement than both a lateral locking plate and a barrel hoop plate, lending support to the contour-matching hypothesis advanced here. The present study extends this evidence by providing a direct biomechanical comparison among three differently sized and designed posterior buttress constructs under standardized loading conditions.

The findings of this study have several clinically relevant implications. First, all three constructs maintained displacement below 2 mm up to 1500 N, supporting the use of posterior buttress plating as an effective fixation strategy for posterolateral tibial plateau fractures, consistent with previous literature [1, 2, 4, 12]. The 1500 N load approximates the forces experienced during early protected weight-bearing in the postoperative period, suggesting that each tested implant may be biomechanically sufficient for this phase of rehabilitation [10, 13].

Second, the comparatively lower displacement of the 2.7 mm distal radius T-oblique LCP under low-to-moderate loading suggests that anatomical congruence between the implant and the posterolateral fragment may be an important determinant of construct stability, potentially to a greater extent than screw diameter alone. This consideration may be particularly relevant given the irregular contour of the posterior tibial surface [6]. In such cases, selecting a plate whose contour closely approximates the three-dimensional morphology of the posterolateral fragment may facilitate optimal implant seating and load distribution while potentially reducing the risk of neurovascular injury during the surgical approach. Implant selection for posterolateral tibial plateau fixation may therefore be guided primarily by anatomical compatibility between plate geometry and fragment morphology rather than by screw size alone. Nevertheless, all three implants demonstrated similar load-to-failure values, indicating that each construct is capable of providing adequate fixation strength. The 2.7 mm distal radius T-oblique LCP, the 3.5 mm R-LOCK locking plate, and the 3.5 mm small T-oblique DCP therefore remain biomechanically viable options, and the final implant selection may reasonably be informed by surgeon preference, local availability, and individual patient anatomy.

Limitations

Several limitations should be acknowledged. First, although Synbone 1110 synthetic tibia ensured inter-specimen reproducibility and eliminated biological variability, they do not replicate the anisotropic and viscoelastic properties of cadaveric bone. Second, the standardized osteotomies do not reproduce the comminution and complexity of in vivo fracture patterns, which limits direct clinical extrapolation. Third, the sample size (n = 5 per group) was supported by a priori and post hoc power analyses for the 500 and 1000 N comparisons, but statistical power was limited at 1500 N and load-to-failure, and a Type II error cannot be excluded. Fourth, because each specimen was tested under sequential loading, displacement measurements were not fully independent; this was accounted for by the repeated-measures analysis, although residual within-specimen correlation may have affected the precision of the estimates. Finally, the loading protocol was restricted to monotonic axial compression and load-to-failure testing; cyclic, shear, torsional, and combined loading were not assessed. Future studies incorporating cyclic, shear, torsional, and combined loading protocols, as well as cadaveric specimens, are warranted to validate and extend these results.

Conclusions

Under axial loading using a synthetic tibial plateau fracture model, the 2.7 mm distal radius T-oblique LCP demonstrated comparatively lower vertical displacement at 500 and 1000 N than the 3.5 mm R-LOCK locking plate and the 3.5 mm small T-oblique DCP. All three constructs maintained displacement below the 2 mm failure threshold across the tested physiological loading range, suggesting that each is biomechanically adequate to resist normal physiological loads. These observations should be interpreted as biomechanical findings under controlled in vitro conditions, and further investigation incorporating cyclic, shear, and combined loading is needed before specific clinical recommendations can be made.

Abbreviations

CT

Computed tomography

ORIF

Open reduction and internal fixation

LCP

Locking compression plate

DCP

Dynamic compression plate

SA

Sagittal angle

DD

Diagonal distance

PCD

Posterior cortical distance

ANOVA

Analysis of variance

LSD

Least significant difference

SD

Standard deviation

Authors' contributions

P.P., T.T., and K.L. were substantially involved in the conception and design of the study. T.T. prepared the bone models and performed the fixation procedures. E.W. and N.T. conducted the biomechanical testing. T.T., P.P., and K.L. drafted the manuscript and performed the statistical analysis. P.N., P.P., T.T., and K.L. assisted with the literature search and contributed to the manuscript draft. T.T., K.L., P.N., and P.P. participated in the study design, drafted the manuscript, and provided critical revisions for important intellectual content. All authors have read and agreed to the published version of the manuscript.

Funding

Research and Open Access funding were enabled and organized by the Faculty of Medicine Vajira Hospital, Navamindradhiraj University.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

This study used only synthetic tibial bone models and did not involve human participants, human tissue, or animals; therefore, formal ethical approval and informed consent were not required. The protocol was nevertheless reviewed and registered by the Institutional Review Board of the Faculty of Medicine Vajira Hospital, Navamindradhiraj University (COE 031/2022)(Date of acknowledgement 30/09/2022).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.