1 Introduction

The distal femur is one of the most common anatomical sites for primary and non-primary tumors of bone.¹ Surgical treatment for distal femur tumors often consists of resection and reconstruction with an endoprosthesis. Although various bone graft options are utilized in this location, endoprosthetic replacement is typically preferred given lower rates of infection and non-union.²

Among lower extremity endoprostheses in general, the reported failure rate is 12%.³ The specific modes of failure, described by Henderson et al. and adopted by the International Society of Limb Salvage (ISOLS) in 2011, include soft-tissue compromise (Type 1), aseptic loosening (Type 2), structural failure (Type 3), infection (Type 4), and tumor progression (Type 5).⁴ For distal femur endoprostheses in particular, recent systematic evidence suggests aseptic loosening and infection are two of the most common failure modes (8.8% and 8.5%, respectively).⁵ However, despite improvement in surgical technique over time with perhaps slight implant innovation, complications and revisions after tumor endoprosthetic replacement of the distal femur are not uncommon.^6–16

Given the relatively high revision and complication rates, as well as prolonged patient survival due to improvements in adjuvant therapy, it is important to accurately assess revision to reliably inform long-term expectations. Therefore, this study investigated the implant-related outcomes following distal femur replacement for oncologic conditions over a consecutive 14-year period. We used a competing risk analysis to accurately describe the cumulative incidence of revision, and we discuss the observed modes of failure within our institution.

2 Materials and methods

3 Results

3.2

3.2 Revision rates

The all-cause revision rate was 21.1% (n = 12) at a mean 65.3 ± 47.3 months due to infection (n = 3, Type 4), structural complications (n = 3, Type 3), aseptic loosening (n = 3, Type 2), and patellar resurfacing (n = 3). Between revised and unrevised cases, there was no difference in age (27.6 ± 17.6 vs. 40.9 ± 21.9 years; p = 0.058), BMI (23.9 ± 4.0 vs. 26.1 ± 7.9 kg/m²; p = 0.389), femoral stem length (16.6 ± 10.7 vs. 15.5 ± 22.7 cm; p = 0.140), stem diameter (12.1 ± 1.8 vs. 11.9 ± 1.7 mm; p = 0.794), or extramedullary prosthesis length (15.9 ± 4.7 vs. 18.8 ± 6.4 cm; p = 0.171) (Table 2).

Table 2 Characteristics among revised cases and those without a revision by last follow-up.

Variable	Revision (n = 12)	No Revision (n = 45)	p - valueb
Age (years)a	27.6 ± 17.9	40.9 ± 21.9	0.058
Body mass index (kg/m2)a	23.9 ± 4.0	26.1 ± 7.9	0.389
Femoral Stem Fixation			0.281
Cemented	10 (83.3)	42 (93.3)
Uncemented	2 (16.7)	3 (6.7)
Femoral Stem Diameter (mm)a	12.1 ± 1.8	11.9 ± 1.7	0.794
Femoral Stem Length (cm)a	16.6 ± 1.1	15.5 ± 2.3	0.140
Extramedullary Prosthesis (cm)a	15.9 ± 4.7	18.8 ± 6.4	0.171
Femoral resection (cm)a	14.8 ± 3.1	16.9 ± 3.9	0.105
Radiation Therapya			0.511
Yes	2 (16.7)	12 (26.7)
No	9 (75.0)	31 (68.9)
Chemotherapya			0.278
Yes	7 (58.3)	12 (26.7)
No	9 (75.0)	31 (68.9)
Follow-up (months)a	94.9 ± 49.5	32.6 ± 32.3	<0.001*

a Mean (standard deviation).

b Chi-Squared Test of Independence or Independent Two-Sample Means t-Test.

Competing risk analysis revealed an estimated cumulative incidence of revision of 12.0% (95% CI, 3.6–25.9%) at five years and 36.5% (95% CI, 12.8–61.0%) at ten years (Fig. 1). Patellar revisions (n = 3) were resurfacing due to pain or maltracking. All three structural failures (Type 3) were due to fracture of the femoral stem at a median 75 months (range, 67–133 months). These patients were revised with a second stem (n = 2) or allograft-prosthetic composite (n = 1). Failure due to aseptic loosening (n = 3, Type 2) occurred in two cemented femoral stems and one press-fit stem at a cumulative median 121 months (range, 15–148 months). These patients were revised to an allograft-prosthetic composite (n = 1) and short-segment fixation device (n = 1), or cemented stem from a press-fit stem (n = 1) (Fig. 2). Failure from infection (n = 3, Type 4) occurred at a median 36 months (range, 3–76 months). The three infected cases underwent hip disarticulation (noted above), resection arthrodesis with a rod and antibiotic cement, and deep excisional-type irrigation and debridement with bushing exchange, respectively.

Close

Fig. 1 Cumulative incidence function using a competing risk analysis with death of the patient as the competing risk. A. The risk of implant revision was 12.0% (95% CI, 3.6–25.9%) at five years and 36.5% (95% CI, 12.8–61.0%) at ten years. B. However, the cumulative mortality incidence was lower at ten years (21.2%, 95% CI, 6.1–42.3%), which was lower than the revision risk for any cause and suggests patients typically undergo revision in their lifetime.

Export to PPT

Close

Fig. 2 Of the three cases revised for aseptic loosening, two were cemented stems, while the third was a press-fit stem. The time to failure among all three ranged from 15 to 148 months. A-B. One cemented stem was significantly loosened (circle) with distal radiolucency and was revised to a short-segment compress fixation device. C-D. The other cemented stem loosened (circle) with evidence of significant radiolucency (arrow) and was converted to an allograft-prosthetic composite. The remaining press-fit stem (not pictured) was revised to a cemented femoral stem at 15 months due to pain and suspected loosening.

Export to PPT

3.3

3.3 Radiographic outcomes

Two of the three cases revised for aseptic loosening had evidence of a radiolucent area of bone greater than 50% of the femoral stem length prior to revision (Fig. 2). The third case (press-fit stem) was revised at 15 months for pain and suspected loosening and was diagnosed intraoperatively. None of the aseptic loosening cases had significant cortical expansion remodeling. Among all the unrevised cases (n = 45), 15.8% (n = 9) had evidence of cortical expansion remodeling at a mean 82.6 ± 44.4 months following surgery (Fig. 3). The incidence of stress shielding among all unrevised cases was 3.5% (n = 2) at a mean 45.5 ± 13.4 months (Fig. 3).

Close

Fig. 3 Patients who undergo distal femoral endoprosthetic replacement should be followed with long-term radiographs to assess for worrisome or reassuring features. A-B. Cortical expansion remodeling (circle), defined as an outgrowth of bone at the proximal tip of the femoral stem, may be reassuring and was identifiable on 15.8% of final radiographs at a mean 82.6 ± 44.4 months. C-D. In a separate endoprosthetic system, stress shielding (arrow) was evident at five years and together with radiolucency (osteolysis) may suggest a constellation of worrisome features. However, this patient did not progress to failure.

Export to PPT

4 Discussion

The present study recorded an all-cause revision rate of 21.1% at 65 months, with the most common causes being infection, structural failure, and aseptic loosening. Modular endoprosthetic replacement is often the treatment of choice for bony malignancies around the knee given the relative ease of reconstruction. Earlier systems were largely custom-designed with a fixed knee mechanism, though were fraught with complications such as articulating component failure.¹⁹ Modern improvements such as modularity and a rotating-hinge platform have likely provided greater longevity, and recent systematic evidence has shown improved survival using these designs.⁵ For stem fixation, however, the data are less clear and are debated among the literature, and this area of research remains actively investigated.^20–23 In our study, the majority of the endoprosthetic systems were cemented (91.2%), and each replacement was modular with a rotating hinged knee.

In their systematic review, Haijie et al. note five-, ten-, and 20-year pooled distal femoral implant survivorships of 78.3%, 70.1%, and 38.3%, respectively.⁵ Therefore, it appears revision is considerably higher in the long-term. Our data also highlight an increased long-term risk of revision, though this may have been confounded by longer follow-up of revised patients. Nonetheless, we a competing risk analysis to estimate risk of revision, and this method has been suggested to be more precise than conventional Kaplan-Meier for modular tumor endoprostheses because it includes patient death as a competing risk.²⁴ The cumulative incidence of revision in our cohort was 12.0% at five years and 36.5% at ten years for any cause (Type 1–5 and patellar revisions). Our data align with Staats et al., who found a similar cumulative incidence of revision of 28.3% at five years and 35.7% at ten years.²⁵ Although our sample size is smaller than Staats et al. which included 183 oncologic patients in their study, the incidence of revision in the current study, especially in the short to midterm, appears to be lower than historical controls.

Among all causes for revision after oncologic distal femur replacement, infection and aseptic loosening are two of the most commonly recorded.¹¹ Haijie et al. report pooled estimates of 8.5% and 8.8% for infection and loosening in their systematic review, respectively.⁵ In our study, the rate of aseptic loosening was 5.3% among all procedures, which is a slight improvement compared to estimates from the literature. The infection rate in the current study was also 5.3% which was slightly improved compared to the literature. We believe a lower rate could potentially be attributed to a combination of improved surgical technique and modern protocols. We did not include such an analysis in the current study, though hypothesized that surgical innovation such as routine betadine and tranexamic acid use, and universal sterility protocols such as laminar flow theaters and surgical hoods may affect infection risk. Ultimately future research is needed to explore these factors in tumor endoprosthetic surgery.

Of note, we did qualitatively describe the radiographic outcomes in all unrevised cases, as well as in each of the three patients with aseptic loosening requiring revision. Piakong et al. utilized a similar radiographic assessment and found that 66% of patients who progressed to aseptic loosening had a radiolucent area of 20% or greater the length of the femoral stem, without cortical expansion remodeling.²⁶ In our study, we noticed that each of the cemented endoprostheses revised for aseptic loosening consistently had a large radiolucent area (>50% of femoral stem), as well as an absence of significant cortical expansion remodeling at the proximal tip of the stem. While the remaining revision of a loosened press-fit stem did not have these features, it was revised at 15 months for pain and therefore may have been too early to identify any worrisome radiographic signs.

Second, 15.8% of all unrevised cases in the current study showed evidence of cortical expansion at the proximal femoral stem tip. This finding implies that implants which tend to survive may display proximal cortical expansion, while those that progress to aseptic loosening may not. Thus, it may be necessary to routinely assess for these findings on long-term radiographs to inform expectations. Ultimately, however, further research is needed to support this hypothesis as well, though the results are important nonetheless and propose radiographic features that might reliably predict aseptic loosening and therefore a revision surgery.

5 Conclusions

Similar to the literature, our results indicate that the revision risk after tumor endoprosthetic replacement of the distal femur increases over time. However, the risk of revision in our institution was lower than historical controls, especially when using a competing risk analysis which has been shown to be more accurate than Kaplan-Meier for outcomes of tumor endoprostheses. Two of the most devastating complications in our group were infection and aseptic loosening, which were slightly improved upon estimates from recent systematic reviews. Unique to the current study is a description of radiographic features that may help identify individuals at risk for revision due to aseptic loosening, or conversely patients with a durable implant. Ultimately, the clinical utility of these findings still needs to be determined, and prevalent issues such as aseptic loosening still need to be investigated in order to determine the best methods of fixation and perhaps any innovation to surgical techniques or implants that could reduce the incidence of this failure mode.

Ethics approval and consent to participate

Rush University Medical Center obtained individual Institutional Review Board approval with an approved waiver of consent prior to beginning any research efforts.

Availability of data

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Funding

No funding sources were required for this study.

Disclosures

ATB: (BMJ Case Reports: Editorial or governing board; Clinical Orthopaedics and Related Research: Editorial or governing board; exparel/pacira: Stock or stock Options; Journal of Oncology Practice: Editorial or governing board; Journal of Surgical Oncology: ad hoc reviewer; Lancet - Oncology: Editorial or governing board; Musculoskeletal Tumor Society: Board or committee member; Onkos Surgical: Paid consultant; Pediatric Blood and Cancer: Editorial or governing board; Rare Tumors: Editorial or governing board; Rush Orthopedic Journal: Editorial or governing board; Swim Across America Cancer Research Grant: Research support); SG: (Onkos Surgical: Paid consultant; Stock or stock Options; USMI: Stock or stock Options); MWC: (Alphatec Spine: IP royalties; Paid consultant; AO Spine North America: Board or committee member); Research support; Cervical Spine Research Society: Board or committee member; CSRS: Research support; DePuy, A Johnson & Johnson Company: Paid presenter or speaker; K2M: Paid presenter or speaker; Musculoskeletal Tumor Society: Board or committee member; North American Spine Society: Board or committee member; Orthofix, Inc.: Paid presenter or speaker; Spinal Elements: Paid consultant. All other authors have no pertinent financial disclosures or pertinent conflicts of interest.

CRediT authorship contribution statement

Charles A. Gusho: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Joshua A. Greenspoon: Formal analysis, Writing – original draft, Writing – review & editing. Bishir Clayton: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Jonathan Bauer: Data curation, Writing – original draft, Writing – review & editing. Matthew W. Colman: Investigation, Supervision. Steven Gitelis: Investigation, Supervision. Alan T. Blank: Conceptualization, Investigation, Methodology, Supervision, Writing – review & editing.