1 Introduction

Proximal femoral replacement (PFR) reconstructed with a modular endoprosthesis allows for early restoration of stability and weight-bearing in the setting of extensive loss of peri- and subtrochanteric bone.^1,2 This is commonly performed in the setting of primary or secondary sarcoma or for failed total hip arthroplasty (THA) complicated by periprosthetic fracture, infection, or prosthesis failure with poor proximal bone stock. Wide margins obtained with extensive resection for oncologic indications may contribute to improved long-term oncologic outcomes and a lower incidence of local recurrence, while PFR for failed THA is typically reserved as a salvage procedure preserving the patient's functional status. However, the extensive soft tissue dissection and resection of proximal bone and soft tissue in both cases is known to contribute to poor tissue tension, abductor insufficiency, and resultant instability.³

Bearings for proximal femoral reconstruction implants include a hemiarthroplasty with a bipolar modular head or the use of a constrained liner to provide additional stability. Prior studies have demonstrated that bipolar hemiarthroplasty carries a decreased risk of dislocation when compared to the use of a constrained liner due to decreased soft-tissue disruption and the large prosthetic head, which confers an increased head-neck ratio and jump distance.^4–6 However, these implant designs are associated with a high revision rate. Although the constrained liner design decreases friction, thereby mitigating wear on the native bony anatomy, these implants are limited by a restricted range of motion. This is the result of impingement of the femoral component on the liner during wide-ranging motions, leading to prosthesis wear and a high rate of dislocation ranging from 4 % to 35 % following PFR with constrained liner.^7,8

The dual mobility construct may lower the rate of dislocation following PFR by combining the larger head articulation of a bipolar hemiarthroplasty with the lower friction concept of the constrained liner.^9,10 The polyethylene liner between the prosthetic head and external metal shell creates two bearings thereby increasing the effective size of the femoral head.^11–13 Moreover, the internal bearing between the large femoral head and polyethylene liner allows for increased range of motion in all directions and further reinforces the stability of the construct.^14,15 While existing literature has described its successful application in PFRs for both neoplastic and non-neoplastic indications, data comparing the outcomes of dual mobility on stability to traditional constructs remains limited.¹⁶

The objectives of this study include: i) retrospectively evaluate three cohorts of patients who underwent resection of the proximal femur for neoplastic and non-neoplastic indications with bipolar, constrained, or dual mobility PFR reconstruction; ii) compare perioperative profiles between the three groups; iii) evaluate the effect of the type of construct on short- and long-term stability by comparing the incidence and time to dislocation.

2 Materials and methods

3 Results

A total of 106 patients were included for analysis, with 46 receiving the bipolar construct, 42 receiving the constrained liner construct, and 18 receiving the dual mobility construct. The three cohorts did not significantly differ in baseline demographic characteristics, comorbid conditions, or perioperative factors (Table 1), with the exception of a lower mean BMI in the bipolar cohort (P = 0.036) and a higher proportion of smokers in the constrained liner cohort (P = 0.002). There was no difference in follow-up time for the cohorts (P = 0.101). Indication for PFRs differed significantly within the three construct types (P < 0.001). The constrained liner construct was primarily utilized in patients with THA failure (76.2 % of constrained liners), whereas the bipolar construct was more readily utilized for the indications of primary sarcoma (45.7 % of bipolar constructs) and metastatic bone disease (50.0 % of bipolar constructs). The dual mobility construct was utilized primarily for metastatic bone disease (44.4 % of dual mobility constructs), followed by THA failure (33.3 %) and primary sarcoma (22.2 %).

Failure by postoperative dislocation was analyzed between the three cohorts (Table 2). The rate of postoperative dislocation did not significantly differ between construct types (bipolar 8.7 % vs. constrained liner 19.0 % vs dual mobility 16.7 %; P = 0.451). However, the mean time to dislocation significantly differed between the groups, with the dual mobility construct (10.68 ± 7.66 months) dislocating significantly later than the bipolar and constrained liners (2.63 ± 2.49 months and 1.87 ± 1.41 months, respectively) (P = 0.009). The mean follow-up time for patients without dislocation was not significantly different between the three liner cohorts (P = 0.131).

Kaplan-Meier Survival Analysis by construct is shown in Fig. 1. Log-rank analysis did not demonstrate significant differences in survival between dual mobility and traditional constructs (P = 0.290).

Close

Fig. 1 Time to Dislocation across Bearing Type. Kaplan-Meier Survival Analysis comparing dislocation survival between liner cohorts (p = 0.29).

Export to PPT

4 Discussion

The present study is the first of its kind to compare rates of dislocation amongst proximal femoral replacement bearing type. Here, we evaluated 106 patients who underwent PFR performed at our institution over a 15-year span using either a bipolar prosthesis, constrained liner bearing, or dual mobility implant. While we found no significant difference in rates of dislocations between the three groups, there was a significant difference in time to dislocation favoring dual mobility implants (P = 0.009). Furthermore, the indication for PFR differed significantly between the three groups, with a preference for constrained liners in THA failure and bipolar constructs in cases of malignancy. Dual mobility seemed to be utilized similarly amongst all indications.

Traditionally viewed as a salvage procedure, proximal femoral replacement is more commonly used in the context of malignancy and revision THA.^2,17 Its use, however, has increased as limb salvage therapy has become more widely accepted and as rates of THA and revision THA continue to increase for aseptic loosening, periprosthetic infection, periprosthetic fracture and non-union. While PFR demonstrates reliable pain relief, good functional outcomes and implant survivorship,^5,18,19 dislocation remains the most common complication.^20,21 Similar to prior studies, the majority of dislocations seen in our study occurred within the first 3 months.²² Rates of dislocation for revision THA vary widely throughout the literature from 8 % to as high as 50 %,^23,24 with slightly higher rates of dislocation noted for PFR.²⁵ This is likely due to excessive resection of soft tissues in cases of malignancy and infection, and loss of abductor musculature in revision arthroplasty.

The use of larger femoral heads and modified bearings in bipolar implants, constrained liners, and dual mobility constructs helps to minimize the risk of dislocation. Shapiro et al. reported no dislocations in 97.6 % of 85 hips revised with a constrained liner for instability and Goetz et al. reported a similar rate of 96 % without dislocations at 3 years in their study of 56 patients revised with a constrained liner.^26,27 Studies using constrained liners in PFR, however, have demonstrated a much higher rate of dislocation ranging from 4 % to 35 %, which is consistent with our study demonstrating a dislocation rate of 19.5 %.^7,8 The increased dislocation rate may be due to the lack of soft tissue tensioning and deficient abductor mechanism that normally provides the necessary stability.

Bipolar hemiarthroplasty has limited use in the context of revision THA unlike its role for musculoskeletal neoplasms. The larger prosthetic head afforded by the bipolar implant increases the jump distance and decreases the overall rate of dislocation with some studies demonstrating dislocation rates as low as 0.9 %.²⁸ The concern remains for acetabular wear and the need for early revision arthroplasty. However, Bernthal et al. reported a 93 % survivorship at 5-years,²⁹ which is similar to other studies showing a 10-year incidence of conversion to a THA of 5 % for malignant disease.³⁰ Studies with longer follow up have shown a significant increase in acetabular wear and the need for revision after 10 years, which suggests a role for bipolar hemiarthroplasty in individuals treated for metastatic disease or patients with high-grade localized disease where prognosis is poor because implant survivorship will often outlast the patient's prognosis. This is consistent with our findings in which bipolar implants were the preferred implant for both primary sarcoma (P < 0.01) and metastatic bone disease (P < 0.05) patients.

Dual mobility implants were only FDA-approved in 2009 in the United States but have quickly become popularized in patients with high risk for instability. A recent 2020 study found that dual mobility usage has increased from 19.5 % in 2012 to 30.6 % in 2018 for revision THA.³¹ Long-term studies are limited, but available literature suggests a survivorship ranging from 99.8 % at 3 years to 98 % at 8.5 years, and a dislocation rate of approximately 1 %.^32,33 While there are no studies to date examining survivorship and revision arthroplasty for dual mobility implants in proximal femoral replacement, it is safe to assume there would be a higher rate of dislocation as observed in trends with bipolar hemiarthroplasty and constrained liners. However, our study demonstrated a higher dislocation rate of 16.7 %. While the smaller cohort size of 18 patients compared to 46 in the bipolar cohort and 42 in the constrained liner cohort likely contributed, the individual circumstances that lead to dislocation should also be taken into consideration.

One of the three dual mobility patients had a dual mobility implant placed in the setting of an internal hemipelvectomy for resection of a metastatic pheochromocytoma after a previously failed THA with a structural allograft. The amount of soft tissue resection and lack of muscle attachment made this patient a high risk for dislocation. Furthermore, her dislocation occurred over 1 year out from surgery making this an atypical timing for dislocation. The second dual mobility patient had a history of metastatic breast carcinoma to the peritrochanteric region complicated by a pathological femur fracture that was treated with a cephalomedullary nail and radiation therapy. The tumor burden progressed, and she developed a non-union thus prompting revision to a PFR. Irradiation of the bone and local soft tissue lead to poor soft tissue healing and integrity, which likely contributed to her persistent instability and recurrent dislocations within 2-months. The final dual mobility implant was placed for the indication of THA failure secondary to periprosthetic fracture in the context of chronic osteomyelitis. Her postoperative course was complicated by an aneurysm of the profunda femoris artery that led to recurrent hematoma formation requiring serial irrigation and debridement of the surrounding soft tissues. Her dislocation occurred nearly 1.5 years after the PFR surgery, which is well outside the typical 3-month window traditionally observed. Both the second and third patient required additional revision surgery for recurrent instability.

Nevertheless, even in the setting of compromised stability due to postoperative complications and individual patient circumstances, there was a significantly greater time from surgery to dislocation in the dual mobility cohort (P = 0.009). This suggests that dual mobility implants provide stability for a longer period of time and are less likely to be complicated by early dislocation than the traditional constructs of PFR. This allows added time for soft tissue healing and scarring to reduce joint hypermobility and risk for dislocation.

The shorter follow up for dual mobility implants creates uncertainty about its long-term complications. Concerns for accelerated wear due to the dual articulation interfaces is the most cited reason for concern.³¹ Osteolysis rates in the older models ranged from 0 to 20 % but use of highly cross-linked polyethylene in the newer models demonstrate much lower rates of wear.^34,35 Our cohort follow up for dual mobility implants was 15.38 ± 15.67 months and was too short to evaluate for any relevant data on accelerated wear and osteolysis. While long-term follow up studies are essential, the use of dual mobility implants in proximal femoral replacement is probably less impactful because it is being utilized in the context of revision arthroplasty and as a salvage procedure with poor proximal bone stock where alternative options are significantly limited.

Based on the current literature and our current findings, we propose that bipolar hemiarthroplasty implants should be used for tumor patients with limited prognosis and/or for metastatic disease since the lower dislocation rate and strong 10-year survivorship makes it an ideal implant with low risk for early revision. The implant choice for revision THA is less clear and may be more surgeon and institution specific. There are currently no available studies comparing constrained liners and dual mobility implants in proximal femoral replacement. However, Eecke et al. recently reviewed data from 46 patients, 15 with constrained liners and 31 with dual mobility, in the context of revision THA and demonstrated higher survival rates, lower dislocation rates and lower acetabular loosening rates in dual mobility implants.³⁶ Until more studies are published that directly compare these two implant designs, there is no clear preference for PFR bearing selection in revision THA.

Limitations of this study include the inherent limitations of a retrospective study. The lack of long-term follow up for each cohort and the small cohort size secondary to the unique surgical procedure being studied also limits extrapolation of the data. Furthermore, our study compiled data from multiple surgeons over an 18-year period, so study protocols were inconsistent amongst all patients reviewed. There were likely changes in technique, implant selection, and surgeon skill over the years. For example, in a retrospective cohort it is difficult to standardize the handling of the abductor-vastus lateralis complex, the integrity and reconstruction of which is critical to implant stability. One area of interest moving forward would be to evaluate patient reported outcomes in order to determine what trends or preferences are noted amongst the three bearing types. It would also be interesting to evaluate whether a dual mobility cohort size similar in size to the other two cohorts would affect significance of dislocation rates.

5 Conclusion

While multiple studies have described superior results with dual mobility for THA and revision THA, its role in PFR remains to be described by literature. This study found similar rates of dislocation following PFR among patients indicated for tumor or failed THA regardless of the construct used. However, patients with a dual mobility prosthesis were less likely to dislocate early within the first three months postoperatively and were more likely to experience longer implant survivorship prior to failure compared to the bipolar hemiarthroplasty and constrained liner cohorts. Additional large-scale longitudinal studies assessing dislocation rates among these unique constructs may represent a valuable area for future analysis.

Consent statement

No consent was needed for this study.

Ethical statement

All procedures were performed in compliance with relevant laws and institutional guidelines, including the Declaration of Helsinki, and have been approved by the appropriate institutional committee(s). This is a retrospective study, therefore IRB approval was obtained at our institution (ORA#20073108).

Funding declaration

There were no funding sources involved in the creation of this study/manuscript.

CRediT authorship contribution statement

Athan G. Zavras: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing, Visualization. Rajko S. Vucicevic: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing, Visualization. Michael P. Fice: Conceptualization, Methodology, Writing – review & editing, Supervision, Validation. Austin Yu: Data curation, Writing – review & editing. Zeeshan A. Khan: Formal analysis, Writing – review & editing. Navya Dandu: Data curation, Writing – review & editing. Alan T. Blank: Conceptualization, Methodology, Supervision. Steven Gitelis: Conceptualization, Methodology, Supervision. Brett R. Levine: Conceptualization, Methodology, Supervision. Matthew W. Colman: Conceptualization, Methodology, Supervision, Project administration.