1 Introduction

Periprosthetic fractures following total knee arthroplasty (TKA) are associated with increased morbidity and mortality due to high rates of nonunion and fracture-related complications.^1–7 The most common fracture site following TKA is the distal femur.^2,4,7,8 Periprosthetic distal femur fractures (pDFF) have an incidence of 0.2–3.5% and occur most commonly in elderly patients and women with bone poor quality.^4,7–9 Recently, the rate of periprosthetic fracture incidence has increased due to a growing elderly population and an increased number of knee arthroplasties.^3,4,7,10 As the number of knee arthroplasties continues to rise, the incidence of periprosthetic fractures will also continue to increase.

Fractures of the distal femur continue to be difficult injuries to treat successfully with reported nonunion rates of 0–22.2%.^11–14 The volume of literature regarding distal femur fracture outcomes and fixation techniques in native knees has increased.^6,15 However, information to guide surgeons and patients on distal femur periprosthetic fractures remains scarce compared to the literature on distal femur fractures in native knees. Surgical fixation techniques for DFF have evolved significantly in the past 10 years especially with the recent rise in utilization of the dual implant technique.^15–17 Dual implant fixation has been associated with a lower nonunion rate and potentially a shorter time to full weight bearing.^16,17 It remains unclear if this new technique has been adopted in periprosthetic femur fractures.

The purpose of this study was to compare outcomes and fixation techniques of distal femur fractures between patients with native knees and those with an ipsilateral total knee arthroplasty. The study results will provide valuable information to counsel patients with pDFF on prognosis and to inform surgeons with new surgical techniques that can potentially improve union rates for pDFF.

2 Material and methods

Institutional review board approval was obtained. A retrospective review was completed on patients with DFF treated with surgical fixation from January 2019–March 2021 at a level one trauma center. Patients were identified utilizing CPT codes 27511 and 27513, open reduction internal fixation (ORIF) of a supracondylar fracture of the femur with or without intra articular extension, respectively. 238 patients were initially included. Patient radiographs and Computed Tomography (when available) were reviewed and all fractures not involving the distal third of the femur based on AO/OTA classification were excluded.¹⁸ Additional exclusion criteria were defined as patient age <50 years, patients with polytrauma, non-ambulatory patients at baseline, revision surgeries, and patients with <90 days of follow-up, Fig. 1. 140 fractures fit the inclusion criteria. A chart review was completed to collect patient demographics including age, gender, Body Mass Index (BMI), smoking status, American Society of Anesthesiology (ASA) classification, AO/OTA fracture classification, presence of TKA, surgical fixation method, mean time to full weight bearing, and complication rate. Fixation method was defined as single (plate or nail) or dual implant (plate + plate or plate + nail) fixation. Complications were defined as re-admission for deep infection, other re-admission, thromboembolic events, and nonunion at 3 months. Statistical analysis comparing the TKA vs native knee cohorts were performed using t-test, chi-square and Fisher's exact test where appropriate (SAS, Cary, North Carolina; GraphPad, San Diego, California). Statistical significance was set at α = 0.05.

Close

Fig. 1 Flowchart exemplifying cohort inclusion methodology. CPT = Current Procedural Terminology DFF = Distal femur fractures TKA = Total knee arthroplasty; N= Number No color is necessary for Fig. 1.

Export to PPT

3 Results

138 patients and 140 fractures with a mean age of 73.5 years (range 50–97) were included in the final analysis. 69 patients (49.3%) with a mean age of 73.9 years were included in the TKA group and 71 patients (50.7%) with a mean age of 73.0 years comprised the native knee group (Table 1). The mean follow up was 433 days (range 90–1225 days). Fracture classification of 33A was observed in 39 native knees vs. 69 in the TKA group, class 33B was observed in 1 NK vs. 0 in the TKA group, and class 33C observed in 31 NK vs. 0 in the TKA group. No other statistical significance in patient demographics between the native knee and TKA cohorts was observed, p > 0.05 (Table 1).

When comparing mean time to full weight bearing, no significant difference was present between the native knee and TKA cohorts, 8.5 vs 8.6 weeks respectively, p = 0.64 (Table 2). Dual implant fixation occurred less frequently in the TKA group, with utilization occurring in 3 patients (4.3%) compared to 11 patients (15.5%) in the native knee group, p = 0.02 (Table 2). There was no difference in complications between the two cohorts: 5 patients (7.2%) in the TKA group vs. the 12 patients (16.9%) in NK group, p = 0.21 (Table 2). There was 1 nonunion reported in the TKA group compared to 6 nonunions that occurred in the native knee cohort (Table 2). No patients with dual implant fixation experienced a nonunion in our cohort.

4 Discussion

Successful treatment of pDFF requires careful attention and continues to be challenging despite increased rates of occurrence and adequate management.^2,3,14 Nevertheless, surgical treatment is favored over nonoperative management due to high rates of nonunion, malunion, and complications associated with conservative techniques.^1,3,19,20 When comparing patient outcomes in this study, there were several similarities between cohorts. Our study found a similar mean time to weight bearing (MTWB) in patients with native knees and TKA, 8.5 and 8.6 weeks respectively, which is unique from available literature. Von Keudel et al. discussed the MTWB following DFF in native knees normally occurred at 6–12 weeks, compared to multiple small studies that reported a MTWB in patients with TKA occurred at 16–20 weeks.^4,7,²¹ This result is likely due to recent literature-supported attempts to decrease the time to weight bearing in patients with distal femur fractures post-operatively, independent of fixation construct (plate, nail, or dual implant) or fracture characteristics.^22,23 The other similar outcome between cohorts was the complication rate. Our study found a decrease in complication rate in patients with TKA (8%) when compared to native knees (16.9%), although the result was not significant. The similar complication rate between NK and TKA cohorts is consistent with current literature, although data is severely limited. However, our study demonstrated a decrease in complication rate of both cohorts in comparison to the widely reported literature range of 19–35% in NK patients and 19–53% in TKA patients.^3,4,9,10,22 This decrease is likely due to the trauma-fellowship training completed by all surgeons in our group as well as the size of each cohort. Based on these similarities, it is reasonable to apply the known native knee DFF data to patients with a pDFF, including implant technique, union rate, complications, and recovery expectations. However, additional research to further determine the similarities and differences between cohorts should be completed to confirm the results in this study.

There has been a recent trend toward dual implant fixation in the treatment of distal femur fractures to promote increased rates of union and reduce time to weight bearing.^3,11 While the literature regarding outcomes of dual implant in native knees continues to increase, there is little data available that studies the outcomes and complications of dual implant in pDFF. Our study exemplifies a significantly higher dual implant utilization rate in native knees (15.5%) than TKAs (4.3%). It should be noted that the same surgeon group in the study perform ORIF for both the NK and the pDFF groups. This finding is likely due to the limited literature on the use of this technique in pDFF fractures.

There are several limitations to this study, including the retrospective design, minimum 90-day follow-up, and small population size. The retrospective study design allows for confounding variables to be present between populations. This bias was minimized due to similar patient demographics and almost identical cohort sizes. All surgeries were performed by trauma fellowship-trained surgeons, reducing this potential bias. Although follow-up was a minimum of 90 days, the institution utilized in this study serves a wide patient population and geographic area, decreasing the probability that patients would present to alternative hospitals for complications or be lost in follow-up. The pDFF cohort composition was limited because only patients with ORIF of the femur were included, and all patients in the pDFF group had 33A fracture classification. This is likely due to the 33B and 33C fractures requiring a distal femur replacement, which were not included in this study. The study size of 140 fractures is the largest known study to date comparing DFF in patients with native knees to patients with a TKA.²⁴ As periprosthetic fracture frequency continues to increase and outcomes comparing DFF in patients with native knees and TKA become available, increasing the cohort size and study power will be achievable.

5 Conclusions

After analyzing the outcomes and complications of distal femur fractures treated with surgical fixation in patients with native knees vs. total knee arthroplasty, this study demonstrated no difference in mean time to full weight bearing or complication rate between cohorts with similar patient demographics. This result leads to the conclusion that known data regarding native knee DFF should be applied to patients with pDFF. There was a significantly lower utilization rate of dual implant fixation in patients with an ipsilateral TKA compared to the native knee cohort. Similar outcomes and complications between the two groups would suggest that the dual implant technique can be applied for pDFF.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Patient's consent

This study is a retrospective chart review study and the IRB noted that sufficient safeguards would be placed to warrant waiver of consent and authorization according to federal regulations regarding human subjects and privacy (medical record review).

CRediT authorship contribution statement

Nicholas Bostrom: Conceptualization, methodology, project administration, formal analysis, investigation, visualization, Writing- Review & Editing, Supervision. Thomas Z. Paull: Data curation, formal analysis, investigation, Writing- Original Draft, visualization, Writing- Review & Editing, Supervision. Mai P. Nguyen: All authors reviewed the results and approved the final version of the manuscript.