Dual locking plate fixation for periprosthetic tibial fractures With a well-fixed tibial component after total knee arthroplasty: a retrospective case series

1 Introduction

Periprosthetic fractures after total knee arthroplasty (TKA) are uncommon but potentially serious complications. In a large-sample nationwide cohort study of primary TKAs, the 10-year cumulative proportion of periprosthetic knee fractures was 1.7%, whereas tibial fractures accounted for 0.2% (1). Despite their low incidence, periprosthetic tibial fractures (PTFs) are associated with substantial risks of complications and secondary intervention (2). Management is challenging because treatment must account for fracture morphology, bone quality, implant stability, and the relationship between the fracture and the tibial component.

The Felix classification is commonly used to characterize PTFs according to fracture location and tibial component stability. Type I fractures involve the tibial plateau, type II fractures occur adjacent to the tibial component or stem, type III fractures are located distal to the prosthesis, and type IV fractures involve the tibial tubercle. Each type is further classified as subtype A, B, or C, indicating a well-fixed component, a loose component, or an intraoperative fracture, respectively (Figure 1) (3). Treatment selection is primarily determined by implant stability, fracture displacement and morphology, bone stock, and patient-related factors. Stable and minimally displaced fractures may be managed nonoperatively in selected patients. Displaced fractures with a well-fixed tibial component can be treated with internal fixation, whereas fractures associated with component loosening generally require revision arthroplasty (2, 4).

Internal fixation of stable Felix type 2A and 3A fractures can preserve the original tibial component but remains technically demanding. The tibial component or stem may occupy the proximal medullary canal and restrict implant selection and screw trajectories. In a previous series of 16 PTFs treated with minimally invasive locking plate fixation, 14 fractures achieved union, whereas two required secondary procedures. Insufficient proximal screw purchase was associated with fixation failure (5). That study included medial, lateral, and bilateral plate constructs and was not designed to determine whether dual plating was superior to single plating.

Recent studies have included different fracture settings, component stability patterns, and treatment strategies, which limits conclusions regarding a specific fixation construct (4, 6). Direct clinical evidence for dual locking plate fixation in postoperative Felix type 2A and 3A fractures with a well-fixed tibial component remains limited. Therefore, this study aimed to describe the surgical strategy and report the clinical and radiographic outcomes of dual locking plate fixation in this selected patient population. The main outcomes of interest were fracture union and retention of the original tibial component. Additional outcomes included knee function, radiographic alignment, reduction quality, and surgery-related complications. Outcomes were reported for the cohort as a whole, without inferential comparison between Felix subtypes.

2 Methods

2.1 Study design and patients

This retrospective case series was approved by the Institutional Review Board of Luoyang Orthopedic-Traumatological Hospital of Henan Province(Henan Provincial Orthopedic Hospital), Approval No. 2025XJS0013-01. The study was conducted in accordance with the Declaration of Helsinki. The requirement for informed consent was waived because of the retrospective study design.

The medical records and imaging studies of patients treated for postoperative PTFs after primary total knee arthroplasty between June 2022 and December 2024 were reviewed. All included patients had at least 12 months of clinical and radiographic follow-up. Fractures were classified according to the Felix classification (3).

Tibial component stability was assessed preoperatively using anteroposterior and lateral radiographs obtained at presentation and, when available, previous postoperative radiographs. A component was considered well fixed when there was no progressive radiolucent line, migration, subsidence, or change in component position. Computed tomography was performed in nine patients to characterize fracture morphology, define the relationship between the fracture and the tibial tray or stem, and assist preoperative planning; it was not used as the sole criterion for determining component stability. Intraoperatively, the tibial component was directly inspected and manually tested for gross motion. The component was considered stable when no gross motion was detected at the bone–implant or bone–cement interface.

The inclusion criteria were: (1) a postoperative PTF after primary total knee arthroplasty; (2) a Felix type 2A or type 3A fracture; (3) a well-fixed tibial component according to the preoperative imaging and intraoperative criteria described above; (4) treatment with a dual locking-plate construct; and (5) clinical and radiographic follow-up of at least 12 months.

The exclusion criteria were: (1) a Felix type 1, type 2B, type 3B, or type 4 fracture; (2) tibial component loosening or periprosthetic joint infection; (3) a fracture following unicompartmental knee arthroplasty; (4) an intraoperative periprosthetic fracture; and (5) incomplete clinical or radiographic data.

During the study period, all patients with postoperative Felix type 2A or 3A fractures and a well-fixed tibial component were treated with dual locking-plate fixation. One patient who received the same fixation strategy was excluded from the final analysis because follow-up was shorter than 12 months. No patient with the same fracture pattern and a well-fixed tibial component underwent single-plate fixation, revision arthroplasty, or nonoperative treatment during the study period.

Twenty-five patients with PTFs were screened. Twelve were excluded: Felix type 1, n = 1; Felix type 2B, n = 1; Felix type 3B, n = 5; Felix type 4, n = 3; fracture after unicompartmental knee arthroplasty, n = 1; and follow-up shorter than 12 months, n = 1. The final cohort comprised 13 patients, including seven with Felix type 2A fractures and six with Felix type 3A fractures (Figure 2). Felix subtype was retained as a descriptive fracture characteristic and was not treated as an independent comparison group.

2.2 Surgical technique

All procedures were led by a chief or attending orthopedic surgeon experienced in knee arthroplasty and fracture fixation. All patients underwent open reduction and internal fixation using a dual locking-plate construct. The construct was used to provide medial and lateral support and to increase the available fixation points when screw trajectories were restricted by the tibial tray, keel, or stem. Plate length and position were determined according to fracture location, fracture morphology, available bone stock, and prosthetic geometry.

The two patients with open fractures underwent initial surgical debridement and temporary wound management, followed by definitive dual locking-plate fixation after the soft-tissue condition was considered suitable. For Felix type 2A fractures, the plates stabilized the proximal metaphyseal segment around the tibial component. For Felix type 3A fractures, longer plates were used to span the diaphyseal fracture. The medial and lateral locking plates and screws were manufactured by Changzhou Wujin Jinxingda Medical Instrument Co., Ltd. (Changzhou, Jiangsu, China). The number, length, and trajectory of screws were individualized to maximize fixation in the available proximal and distal bone while avoiding interference with the tibial tray, keel, or stem. Locking screws were directed around the prosthetic component, and bicortical purchase was obtained when permitted by the available screw corridors.

All 13 fractures occurred around cemented tibial components. Screw trajectories were planned around the tibial stem and cement mantle to obtain fixation in the available proximal bone while avoiding interference with the tibial component. Fracture reduction, plate position, screw placement, and coronal and sagittal alignment were confirmed using intraoperative fluoroscopy. Soft-tissue stripping was minimized while obtaining direct fracture reduction. No autograft, allograft, or bone-graft substitute was used.

2.3 Postoperative rehabilitation protocol

All patients followed a common postoperative rehabilitation protocol. Active and passive knee range-of-motion exercises were initiated within the first postoperative week when wound condition and pain permitted. Partial weight-bearing with assistive devices was generally started 4–6 weeks after surgery. Weight-bearing progression was individualized according to fixation stability, radiographic healing, pain, and surgeon assessment. Full weight-bearing was permitted after satisfactory clinical and radiographic evidence of fracture union.

2.4 Data collection and clinical evaluation

Variables included age, sex, body mass index, injured side, injury mechanism, Felix subtype, time from primary TKA to fracture,time from fracture to definitive fixation, tibial component fixation method, concomitant injuries, documented history of osteoporosis, recorded anti-osteoporosis treatment, operative time, estimated blood loss, implant system, follow-up duration, and weight-bearing progression.

Clinical outcomes included knee range of motion (ROM), Knee Society Score (KSS) (7), and Hospital for Special Surgery Knee Score (HSS) (8). Assessments were recorded after the acute fracture but before fixation, at 6 months after surgery, and at the final follow-up. The preoperative assessment represented the acute post-fracture condition and was not considered the pre-fracture functional baseline.

ROM was measured using a standard goniometer by a member of the treating orthopedic team during routine clinical assessment. Because of the retrospective study design, the medical records did not consistently document whether active or passive ROM was assessed, the patient position, the anatomical landmarks used for goniometer placement, or whether repeated measurements were performed. Pre-fracture ROM and functional scores were not consistently available and were therefore not analyzed.

2.5 Radiographic evaluation and complication assessment

Radiographic outcomes included fracture union, time to union, mechanical proximal tibial angle (MPTA), posterior tibial slope (PTS), tibial component stability, and reduction quality. Anteroposterior and lateral radiographs were independently evaluated by two orthopedic surgeons who were unaware of each other's measurements. Interobserver reliability for MPTA and PTS was assessed using a two-way random-effects, absolute-agreement, single-measurement intraclass correlation coefficient with 95% confidence intervals. The mean of the two measurements was used for analysis. Disagreements in categorical assessments were resolved by consensus.

Fracture union was defined as bridging callus across at least three cortices on anteroposterior and lateral radiographs, together with the absence of pain or tenderness at the fracture site during weight-bearing (9). Delayed union was defined as lack of radiographic progression by 6 months after surgery. Nonunion was defined as absence of fracture union by 9 months or the need for additional surgery to achieve union (10).

Tibial component stability during follow-up was determined by the absence of progressive radiolucent lines, migration, subsidence, or change in component position. Reduction quality was assessed on early postoperative radiographs according to coronal and sagittal angulation, cortical contact, and residual displacement. Residual displacement was measured using the digital ruler in the picture archiving and communication system. No external calibration marker was used. Reduction was graded as good when angulation was ≤5° and estimated residual displacement was ≤5 mm; acceptable when angulation was >5° to ≤10° or displacement was >5 to ≤10 mm; and poor when either value exceeded these thresholds or revision was required.

Recorded complications included infection, wound problems, loss of reduction, implant failure, screw loosening, delayed union, nonunion, refracture, thromboembolic events, reoperation, and revision arthroplasty.

2.6 Statistical analysis

Statistical analyses were performed using SPSS version 21.0 (IBM Corp., Armonk, NY, USA). Continuous variables were summarized as mean ± standard deviation, and categorical variables were reported as number and percentage. Changes in ROM, KSS, and HSS knee scores across the acute post-fracture preoperative assessment, 6-month follow-up, and final follow-up were evaluated using the Friedman test. When the overall test was significant, pairwise comparisons were performed using the Wilcoxon signed-rank test with Holm adjustment for multiple comparisons. Felix subtype and injury-energy category were retained as descriptive injury characteristics, and no inferential between-subtype or trauma-energy comparisons were performed. All tests were two-sided, and a P value <0.05 was considered statistically significant.

3 Results

3.1 Patient and perioperative characteristics

A total of 13 patients treated between June 2022 and December 2024 were included. Seven patients (53.8%) had Felix type 2A fractures, and six (46.2%) had type 3A fractures. The cohort comprised six men and seven women, with a mean age of 80.4 ± 2.5 years and a mean body mass index of 20.5 ± 0.6 kg/m². Six fractures involved the left side and seven involved the right side. Five injuries (38.5%) resulted from high-energy trauma, and eight (61.5%) resulted from low-energy trauma. Two fractures (15.4%) were open and were managed with initial debridement followed by staged definitive plate fixation. Two patients (15.4%) had concomitant ipsilateral ankle fractures, whereas no concomitant traumatic injuries were recorded for the remaining 11 patients (84.6%). The mean interval from primary TKA to fracture was 7.8 ± 2.4 years, and the mean interval from fracture to definitive fixation was 23.4 ± 12.0 days. The mean operative time was 155.0 ± 19.7 min, and the mean estimated intraoperative blood loss was 265.8 ± 20.2 mL. Partial weight-bearing was initiated at a mean of 5.3 ± 0.8 weeks, and full weight-bearing at a mean of 16.8 ± 0.8 weeks. The mean follow-up duration was 20.5 ± 1.6 months (Table 1).

Table 1. Baseline and injury characteristics of the overall cohort (n = 13).

Values are presented as mean ± SD or n (%), unless otherwise indicated. BMI, body mass index; TKA, total knee arthroplasty.

3.2 Clinical outcomes

At the acute post-fracture preoperative assessment, the mean ROM was 17.5 ± 6.5°, the mean KSS was 26.9 ± 3.8, and the mean HSS knee score was 20.4 ± 2.9. At 6 months after surgery, the corresponding values were 110.7 ± 1.9°, 80.5 ± 2.5, and 70.5 ± 2.8, respectively. At the final follow-up, the mean ROM was 122.9 ± 2.2°, the mean KSS was 89.9 ± 3.0, and the mean HSS knee score was 86.7 ± 2.5 (Table 2).

Table 2. Clinical outcomes over time for the overall cohort (n = 13).

Values are presented as mean ± standard deviation. Overall differences across the three time points were evaluated using the Friedman test. Pairwise comparisons were performed using the Wilcoxon signed-rank test with Holm adjustment; all adjusted pairwise P values were <0.001. The preoperative assessment was performed after the acute fracture and was not considered the pre-fracture functional baseline.

ROM, KSS and HSS knee scores differed significantly across the three assessment time points. Pairwise comparisons with Holm adjustment showed significant differences between the acute post-fracture preoperative assessment and 6 months, between the acute post-fracture preoperative assessment and final follow-up, and between 6 months and final follow-up for all three outcomes (all adjusted P < 0.001).

The post-fracture preoperative measurements reflected the acute injury status and were not considered pre-fracture functional baselines. Accordingly, the observed changes should not be interpreted as effects attributable solely to dual locking-plate fixation.

3.3 Radiographic outcomes

All 13 fractures achieved radiographic union, first documented at a mean of 16.4 ± 1.0 weeks. No delayed union or nonunion was observed. Reduction quality was graded as good in eight patients (61.5%) and acceptable in five (38.5%); no poor reduction was recorded (Table 3).

Table 3. Surgical, rehabilitation, radiographic, and complication outcomes of the overall cohort (n = 13).

Values are presented as mean ± SD or n (%), unless otherwise indicated. ICC values are single-measurement estimates from a two-way random-effects, absolute-agreement model. CI, confidence interval; ICC, intraclass correlation coefficient; MPTA, mechanical proximal tibial angle; PTS, posterior tibial slope.

The mean postoperative mechanical proximal tibial angle was 87.2 ± 1.4°, and the mean postoperative posterior tibial slope was 2.3 ± 0.4°. The original tibial component was retained in all patients, with no recorded loss of reduction, component migration, subsidence, or change in component position during follow-up.

Interobserver reliability was excellent for both MPTA (ICC: 0.93; 95% CI: 0.80–0.98) and PTS (ICC: 0.91; 95% CI: 0.75–0.97).

3.4 Complications

No recorded surgery-related complications occurred during follow-up. There were no cases of superficial or deep infection, wound complication, implant failure, screw loosening, refracture, symptomatic thromboembolic events, reoperation, or revision arthroplasty. Individual rehabilitation and outcome data are provided in Supplementary Table 1.

Representative cases are shown in Figures 3, 4.

4 Discussion

4.1 Principal findings

This retrospective case series included 13 patients with Felix type 2A or 3A PTFs around well-fixed cemented tibial components. All fractures achieved radiographic union after dual locking-plate fixation. The original tibial component was retained in every patient, and no loss of reduction or component migration was observed during follow-up. Knee ROM, KSS, and HSS knee scores increased across the three assessment time points. No fixation failure, reoperation, or revision arthroplasty was recorded.

PTFs remain difficult to manage and are associated with a substantial risk of complications and further surgery. A recent systematic review reported overall complication and reoperation rates of 38% and 28%, respectively (2). Outcomes have varied considerably among published case series. Bauer et al. reported infection, secondary displacement, nonunion, and frequent revision surgery in a mixed cohort of 15 fractures (11). In contrast, Sim et al. reported fracture union without secondary intervention in a multicentre series limited to well-fixed Felix type 2A fractures treated with locking plates (12). These differences probably reflect variation in implant stability, fracture morphology, fixation strategy, and patient selection.

The present cohort was deliberately restricted to fractures around well-fixed components that were treated with the same fixation strategy. This produced a series that was uniform with respect to component stability and fixation strategy, but the findings may not apply to other fracture patterns or loose components. The observed changes in knee function represent recovery from the acute post-fracture condition. They cannot be attributed solely to the dual-plate construct because pre-fracture functional data and an independent comparison group were unavailable. The main contribution of this study is therefore the description of a feasible fixation strategy in selected patients. It does not establish that dual plating is necessary or superior to single-plate fixation, revision arthroplasty, or other treatment options.

4.2 Interpretation of clinical and radiographic outcomes

All fractures achieved radiographic union, and no loss of reduction or tibial component migration was observed during follow-up. The mean union time of 16.4 weeks was within the range reported for surgically treated PTFs (5, 12). However, relevant studies indicate that patients with postoperative PTFs face a high risk of reoperation and tibial component revision, with infection being a common reason for further treatment (6, 13). The results of the present study should therefore be viewed in the context of its small and highly selected cohort.

Retention of the original tibial component was an important finding. Implant stability is considered a central factor in treatment selection for PTFs (4). Recent reviews suggest that displaced fractures around a stable component may be treated with fracture fixation, whereas a loose component usually shifts treatment toward revision arthroplasty (14). All components in the present series were judged to be stable before fixation and were confirmed intraoperatively. This should be taken into account when comparing the results of this study with cohorts comprising mixed cases.

ROM, KSS, and HSS knee scores increased across the three assessment points. The preoperative values were obtained after the acute fracture and did not represent pre-fracture knee function. Acute knee injury can cause arthrogenic muscle inhibition, quadriceps activation failure, and loss of active extension (15). The observed changes therefore reflect recovery from the acute injury through fixation, fracture healing, pain relief, and rehabilitation. They cannot be attributed solely to the dual-plate construct, and the available data do not show whether patients returned to their pre-fracture functional level.

4.3 Rationale and technical considerations for dual locking-plate fixation

Routine dual plating cannot be justified by Felix subtype alone. The fractures in this series were not uniformly comminuted. The main technical concern was the restricted space available for fixation around a cemented tibial component. The tibial tray, stem, and cement mantle limited bicortical screw trajectories and reduced the number of fixation points available in the proximal segment. In a clinical series of PTFs, fixation failure was more frequent when the proximal fragment had fewer than eight cortices of screw purchase. This finding suggests that the amount and distribution of proximal fixation may be more important than fracture comminution alone (5).

In the present series, the medial and lateral plates were used to distribute fixation around the prosthesis. In Felix type 2A fractures, the second plate provided additional fixation around the short metaphyseal segment, where screw placement was constrained by the tibial stem and cement mantle. In type 3A fractures, longer plates were used to span the diaphyseal fracture and provide support on both sides of the tibia. The intended purpose was to reduce dependence on a single plate for coronal alignment and rotational control. A recent biomechanical study found that supplemental medial plating increased the stiffness of a lateral locked-plate construct, although the magnitude of the benefit depended on the configuration of the medial plate (16).

Most recent clinical evidence on dual plating comes from periprosthetic distal femoral fractures and is therefore indirect. In a 2024 retrospective cohort, double plating was selected more often for low fractures and fractures with metaphyseal comminution. Revision-free survival was not significantly different from that observed after single plating or distal femoral arthroplasty (17). A 2025 meta-analysis associated dual plating of distal femoral fractures with higher union rates and lower nonunion rates, but operative time was approximately 27 min longer (18). These findings illustrate the potential mechanical benefit and additional surgical burden of a second plate.

The decision to add a second plate should therefore be based on specific fixation goals. These include obtaining sufficient screw purchase, supporting both sides of the tibia, controlling multiplanar instability, and avoiding interference with the prosthesis. The possible gain in stability must be balanced against greater surgical exposure, longer operative time, additional soft-tissue disturbance, and implant prominence (18, 19). These considerations support the use of supplemental plating when prosthetic geometry limits secure screw placement or when balanced medial and lateral support is required. Because all eligible patients in this series received dual plating, the study cannot determine whether a single plate would have produced similar outcomes.

4.4 Clinical implications, patient selection, and rehabilitation

The findings of this study support the selective use of dual locking-plate fixation in PTFs with a well-fixed component. Treatment planning should first consider tibial component stability, bone stock, fracture morphology, and the feasibility of obtaining secure fixation around the prosthesis. Implant stability and remaining bone stock are central to the choice between fracture fixation and revision arthroplasty. A stable component with adequate surrounding bone can usually be retained, whereas component loosening or insufficient bone stock commonly requires revision reconstruction (14).

Dual plating may be particularly useful when the tibial tray, stem, or cement mantle limits the screw trajectories available to a single plate. A short proximal segment, poor cortical purchase, multiplanar instability, or inadequate support on one side of the tibia may further favor the use of a second plate. In these situations, medial and lateral plates can increase the available fixation points and provide support on both sides of the tibia. Felix subtype remains useful for describing fracture location, but the fixation construct should be determined by the mechanical requirements of the individual fracture rather than by fracture classification alone.

Postoperative rehabilitation should be tailored to both construct stability and patient condition. In this series, knee motion was initiated during the first postoperative week, partial weight-bearing began at a mean of 5.3 weeks, and full weight-bearing began at a mean of 16.8 weeks. Weight-bearing progression after lower-limb fracture fixation should account for fixation stability, bone quality, pain, associated injuries, medical comorbidities, and the patient's ability to comply with rehabilitation (20). Evidence from surgically treated tibial plateau fractures suggests that earlier weight-bearing can be safe in selected patients with stable fixation, although fracture patterns and fixation constructs differ substantially among studies (21). These findings cannot be directly extrapolated to PTFs. In this setting, progression of weight-bearing should remain guided by serial radiographs, clinical symptoms, fixation stability, and associated injuries.

4.5 Limitations and future directions

This study has several limitations. It was a single-center retrospective case series with only 13 patients and no comparison group. The findings therefore cannot establish the effectiveness or superiority of dual plating. The small sample and limited follow-up may also underestimate uncommon or late complications. In addition, all fractures occurred around cemented tibial components, which limits extrapolation to uncemented implants.

Pre-fracture functional data were unavailable, and the preoperative assessments reflected the acute post-fracture condition. ROM measurements were extracted from routine clinical records, without consistently documented protocols for active or passive assessment, patient positioning, anatomical landmarks, or repeated measurements. Standardized technique is important for reproducible knee ROM assessment. Residual displacement was measured without an external calibration marker, so radiographic magnification may have affected linear measurements. Future multicenter studies should use standardized functional and radiographic assessments, longer follow-up, and comparative cohorts to define the indications for dual plating more clearly.

5 Conclusion

In this small retrospective case series, dual locking-plate fixation was followed by fracture union, maintenance of alignment, and retention of the original tibial component in selected Felix type 2A and 3A PTFs with a well-fixed component. This approach may be useful when prosthetic geometry limits screw placement and secure fixation cannot be achieved easily around the tibial component. Larger comparative studies are required to define its indications and determine whether it offers advantages over other treatment strategies.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmscd.2026.1885283/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Luigi Zanna, Santa Maria Annunziata Hospital, Italy

Reviewed by: Apoorva Kabra, Sydney Orthopaedic Research Institute Ltd, Australia

Gregorio Secci, Careggi University Hospital, Italy