Bipolar versus total hip modular prosthesis following proximal femoral resection for primary bone tumors: A comparative study

INTRODUCTION

The proximal femur is a common site for primary bone tumors.^[1] The management of tumors of the proximal femur has evolved, with reconstructive surgery being the main principle to achieve limb salvage and retention of function.^[2]

Improvements in treatment for primary bone neoplasms have resulted in increased long-term survival rates. Many of these patients are young and expect to lead active lives, placing greater demands on their implants.^[3]

One of the controversial debates is whether to use bipolar hemiarthroplasty (BHA) or total hip arthroplasty (THA). Although a study found no acetabular erosion with the bipolar modular prosthesis of the proximal femur, another study’s univariate analysis revealed that younger patients (age ≤50 years) are more likely to undergo revision surgery. The main disadvantage of THA was higher dislocation rates than BHA.^[4,5]

This debate has a historical basis in cases of femoral neck fracture, particularly when considering arthroplasty. Historically, THA was rarely done for displaced femoral neck fractures. However, more recent studies have identified several potential benefits of THA over BHA, including superior functional outcomes, reduced pain, improved ambulation, and lower rates of reoperation. These potential benefits hold promise for the future of orthopedic surgery.^[6]

This study was conducted on patients who underwent wide resection, surgical excision of the proximal femur, and replacement with modular prostheses for primary bone tumors — either BHA or THA — to compare function, complications (mainly dislocation rates) in both groups, and the rate of acetabular erosion in the BHA group.

MATERIALS AND METHODS

A retrospective comparative study was done between January 2017 and October 2020 at two hospitals on patients with proximal femoral primary tumors who underwent wide resection and reconstruction by a modular prosthesis, by the same surgical team.

Exclusion criteria

Patients with proximal femur metastatic disease or who had proximal femoral resection for trauma were excluded from the study.

After revising our inclusion criteria, only 54 patients were included in the statistical analysis: 33 in the bipolar group (BHA) and 21 in the total hip group (THA). Other than those patients, four patients were excluded from the BHA group: two patients were lost to follow-up, and two patients died within 1 year. Five cases were excluded from the THA group: 3 lost to follow-up, and two died within 1 year.

The staging was performed preoperatively, and some patients received neoadjuvant chemotherapy; 22 patients in the BHA group, compared to 8 patients in the THA group. A new radiograph and magnetic resonance imaging (MRI) were obtained preoperatively, and resection was performed using a standard lateral approach.

A longitudinal incision was made through the iliotibial band, followed by detachment of the gluteus maximus insertion. The external rotators and abductors were identified and resected after the sciatic nerve was located and isolated. The vastus lateralis was reflected, and the length of the tumor to be resected was determined. Feeding vessels were ligated, and the femoral vessels, including the profunda femoris vessels, were identified and ligated. The rectus femoris was dissected, followed by circumferential dissection of the capsule with an effort to preserve as much as possible; finally, the head was dislocated. Muscles were resected as needed to ensure clear margins. For the bony margin, an additional 2 cm was resected, and a sample was taken from the distal medulla to provide safety margins in all patients.

We used different prosthetic systems according to availability. The acetabulum was reamed during THA, and the trial cup was placed in an anatomical position, guided by the transverse acetabular ligament [Figure 1].

Cemented prostheses were used for reconstruction in all cases of the BHA group. In the THA group, seven patients received a fully cementless modular prosthesis (33.33%), eight patients had a cementless cup on a cemented stem (38.09%), and six patients underwent reconstruction with a fully cementless modular prosthesis (28.57%).

The abductors were attached to the modular prosthesis using non-absorbable sutures [Figure 2]. If the greater trochanter was salvaged, attachment to the prosthesis was achieved with a stainless steel wire.

Close

Figure 1:

The transverse acetabular ligament guided the reaming (A) and positioning of the cup (B)

Export to PPT

Close

Figure 2:

Attachment of the abductors by non-absorbable sutures and also the rotators.

Export to PPT

Postoperatively, patients received chemotherapy if necessary, based on diagnosis; 25 patients in the BHA group compared to 9 in the THA group. Only one patient in each group required post-operative radiotherapy. Follow-up was conducted clinically at 2 weeks, 6 weeks, and then every 6 weeks for the 1^st year, followed by every 3 months for the 2^nd year, and then every 6 months thereafter until the end of the 5^th year. A computed tomography scan was performed every 3 months for the first 2 years, and a bone scan was conducted every 6 months thereafter. If recurrence was suspected, an MRI was requested.

Functional scores – the Harris Hip score (HHS)^[7] and the Musculoskeletal Tumor Society Score (MSTS)^[8] – were obtained starting from 6 months postoperatively, when possible.

The mean age at the time of surgery was 31.27 years in the BHA group (range 15–60 years), and 27.19 years in the THA group (range 15–58 years) [Table 1]. Seventeen males and 18 females were included in the BHA group, while the THA group consisted of nine males and 12 females.

Pre-operative diagnoses for the BHA group were Ewing’s sarcoma in 12 patients (36.4%), osteosarcoma in 10 patients (30.3%), chondrosarcoma in 6 patients (18.2%), pleomorphic sarcoma in 3 patients (9.1%), fibrosarcoma, and synovial sarcoma for one patient each (3%). For the THA group, pre-operative diagnoses were revision cases in 11 patients (52.4%), Ewing’s sarcoma in 7 patients (33.3%), osteosarcoma, pleomorphic sarcoma, and giant cell tumor in 1 patient each (4.8%). Six patients in the BHA group had pre-operative lung metastasis (18.2%), compared to 2 patients in the THA group (9.5%).

The mean operative time was 3.94 h (range, 2–7 h) in the BHA group and 4.19 h (range, 2–6 h) in the THA group [Table 1]. The mean resection length was 17.88 cm (range 12– 29 cm) in the BHA group, while it was 19.71 cm in the THA group (range 8–39 cm). Blood loss was calculated in units, each unit equal to 500 cc. Estimated blood loss was 2.15 units (range 1–8 units) and 2.57 units (range 1–5 units) for the BHA and the THA groups, respectively. The total follow-up period, calculated in months, was 87.27 months (range, 12– 172 months) in the BHA group compared to 33.95 months (range, 12–117 months) in the THA group [Table 1].

Data were coded and entered using the Statistical Package for the Social Sciences version 26 (IBM Corp., Armonk, NY, USA). Data were summarized using the mean, standard deviation, median, minimum, and maximum for quantitative data, and frequency (count) and relative frequency (percentage) for categorical data.

Comparisons between quantitative variables were done using the non-parametric Mann–Whitney test.^[9] To compare categorical data, the Chi-square test was used. An exact test was used instead when the expected frequency was <5.^[10] Survival curves were plotted by the Kaplan–Meier method and compared using the log-rank test. P < 0.05 were considered statistically significant.^[11]

RESULTS

Prosthesis survival curves were plotted by the Kaplan– Meier method. Prosthesis survival in the BHA group was 97% at 1 year and 83% at 5 years, compared to 100% at 1 year and 90% at 5 years for the THA group [Figure 3]. In the BHA group, five prostheses underwent revision, and two patients underwent hindquarter amputations. In contrast, two patients in the THA group underwent revision. The mean time to prosthesis failure in the BHA group was 65.25 months (range, 1–172 months), and 70.50 months (range, 24–117 months) in the THA group. No statistically significant difference was found between the two groups (P: 1.000).

Functional scoring was done using the HHS and MSTS scores. It was obtained 6 months postoperatively and on every subsequent visit. For statistical analysis, we relied on the 1-year HHS and MSTS scores, as well as the last follow-up, which represents the most recent time we could obtain them, even if it was before 1 year had passed. The MSTS score and the HHS were assessed in only 30 patients in the BHA group at 1 year, whereas both were evaluated in 19 patients in the THA group. The MSTS score and HHS were assessed at the last follow-up for 31 patients in the BHA group, whereas they could be evaluated for all patients in the THA group.

The mean MSTS score at 1 year was higher in the THA group (26.32 points; range, 22–30 points) than in the BHA group (24.10 points; range, 18–29 points), with a statistically significant difference (P: 0.007). The mean MSTS score at the last follow-up was higher in the THA group (25.19 points; range, 18–30 points) than in the BHA group (23.80 points; range, 10–30 points), but the difference was not statistically significant (P: 0.266).

Close

Figure 3:

Kaplan-Meier chart for prosthesis survival. Group A represents the bipolar hip arthroplasty group while Group B represents the total hip arthroplasty group. Log-rank test P-value for the Kaplan–Meier chart was 0.316. FU: Follow-up.

Export to PPT

The mean HHS at 1 year for the THA group, 90.53 points (range, 64–100 points), was statistically significantly better than that of the BHA group, 87.55 points (range, 65–96 points) (P: 0.032). The mean HHS at the last follow-up was higher in the THA group (85.62 points; range, 42–100 points) than in the BHA group (84.60 points; range, 26–100 points), but the difference was not statistically significant (P: 0.686).

Two cases were dislocated in the BHA group (6.1%), while one case was dislocated in the THA group (4.8%) due to a dashboard injury. There was no statistically significant difference in the incidence of dislocation between the groups (P: 1.000). Dislocation occurred at 1 month in the BHA group, whereas it occurred at 4 months in the THA group. Regarding two cases from the BHA group, one patient underwent closed reduction and subsequently experienced no further dislocation. In contrast, the other patient underwent open reduction, but the prosthesis was subsequently removed due to redislocation. The patient in the THA group underwent closed reduction, and no other dislocation occurred.

There was no correlation between resection length and the incidence of dislocation. The mean resection length in the whole dislocated cases in all patients was 19.33 cm (range 16–26 cm) compared to 18.55 cm (range 8–39 cm) in the non-dislocated instances, with no statistically significant difference (P: 0.721). The mean resection length in the dislocated cases of the BHA group was 21.00 cm (range 16–26 cm) compared to 17.68 cm (range 12–29 cm) in the non-dislocated instances, with no statistically significant difference (P: 0.500). The mean resection length in the dislocated cases of the THA group was 16 cm compared to 19.90 cm (range 8–39 cm) in the non-dislocated instances, with no statistically significant difference (P: 0.857).

Acetabular erosion occurred in 9 patients in the BHA group (27.3%) [Figure 4]. The mean time of acetabular erosion was 79.11 months (range 36–172 months). Five patients underwent conversion to a total hip modular prosthesis (55.6%), three were treated conservatively with just follow-up as there was no pain, and the last one underwent hindquarter amputation as recurrence was detected during the operation of conversion to a total hip. There was no relation between the age of the patient and the incidence of acetabular erosion, despite the mean age of patients with acetabular erosion being lower, at 24.76 years (range, 15–56 years), compared to 33.75 years (range, 15–60 years) in patients with no acetabular erosion. This difference was not statistically significant (P: 0.245).

Prosthetic infection occurred in 5 patients in the BHA group (15.2%) and in 1 patient in the THA group (4.8%), with no statistically significant difference (P: 0.386). The mean time of infection in the BHA group was 13.80 months (range 0–60 months), compared to the immediate post-operative period in the THA group.

DISCUSSION

One of the controversial debates following the reconstruction of the proximal femur is whether to use BHA or THA on a proximal femoral modular prosthesis. The debate has intensified with recent advancements in neoadjuvant chemotherapy, which have improved survival rates. Houdek et al. in 2016 reported that younger patients under 50 years old were more likely to undergo revision surgery to convert BHA to THA (8 patients, most suffering from primary bone tumors).^[5] However, a study has shown that there is no incidence of acetabular erosion with bipolar modular prostheses.^[4]

Close

Figure 4:

A female patient aged 15 years underwent proximal femoral resection and replacement for chondrosarcoma of the proximal femur. (A and B) Immediate post-operative radiograph. (C) Acetabular erosion that occurred after 14 years. (D and E) Immediate post-operative radiograph after revision surgery with grafting of the acetabulum. (F and G) 1-year follow-up radiograph with the full incorporation of the graft.

Export to PPT

The reported dislocation rate has influenced the debate over whether to use BHA or THA, which was 3 times higher than that observed in BHA modular prostheses.^[12] In addition, there has been a prevailing belief that THA is associated with a higher rate of blood loss and infection.

This study reports a limb salvage success rate of 90.9% in the BHA group and 100% in the THA group. Previous research has shown that limb salvage success rates range from 92.7% to 100%.^{[1,3-5,13-15]} Regarding prosthesis survival, the 1-year, 5-year, and last follow-up rates in the BHA group are 97%, 78.8%, and 83%, respectively. In contrast, the 1-year and 5-year prosthesis survival rates in the THA group are 100% and 90%, respectively. Past studies have reported prosthesis survival rates ranging from 74% to 97% at 5 years and from 67% to 87.5% at 10 years.^[3,4,13,14]

We evaluated functional outcomes using both the MSTS score and the HHS. Our findings indicate that THA offers better functional results than BHA, with a statistically significant difference seen at 1 year. Mean MSTS scores at 1 year and last follow-up are 24.10 (18–29) and 23.80 points (10–30) in the BHA group, compared to 26.32 (22–30) and 25.19 (18–30) in the THA group, indicating good results. Bernthal et al. reported better outcomes with a BHA modular prosthesis, with a mean score of 26.3 points (13–30).^[13] Houdek et al. found MSTS scores for BHA to be 18 (5–28) and for THA to be 19.2 (8–25).^[5]

Mean HHS at 1 year and at last follow-up are 87.55 (65–96) and 84.60 points (26–100) in the BHA group, compared to 90.53 (64–100) and 85.62 (42–100) in the THA group, with a good result for the BHA group and an excellent result for the THA group. Houdek et al. have shown that the mean HHS for BHA was 81 (44–95), the same as the THA, at 81 (55–93).^[5]

Regarding complications, the primary goal of this study is to compare the incidence of dislocation between THA and BHA Proximal Femoral Modular Prosthesis; there is no significant difference between the two groups. In our research, the dislocation rate for THA (4.8%) is lower than that for BHA (6.1%). This may be due to the larger number of cases in the BHA group, and all the cups in the THA group were placed by arthroplasty surgeons, ensuring proper positioning. Houdek et al. reported an overall dislocation rate of 7% with THA, which was higher than with BHA but not statistically significant.^[5] Menendez et al. and Puchner et al. found that the THA group dislocated 3 times more often than the BHA group, with rates of 17.6% versus 6.4%, and 33% versus 11%, respectively.^[1,12] Calabró et al. reported an overall dislocation rate of 3.95%, with a higher rate in THA (6.25%) compared to BHA (3.2%).^[16] Sokolovski et al. observed a dislocation rate of 11.3% in THA.^[15] Dislocation rates for BHA modular prostheses in various studies ranged from 3% to 23%.^{[3,4,13,14,17]}

The rate of acetabular erosion in this study (27.3%) is higher than what is reported in the literature. This may result from including only patients with primary bone tumors, who generally have better survival rates compared to those with metastatic lesions, and tend to be younger. However, no significant correlation was found between patient age and the presence of acetabular erosion. We have long-term follow-up data for the BHA group, with a mean of 78.72 months (range, 12–172 months), which appears sufficient to evaluate prosthesis longevity. Finstein et al. reported a 0% incidence of acetabular erosion in BHA^[4]; other studies have shown low rates, from 1% to 5.8%.^[5,13,14,16] Donati et al.^[18] reported a 12% incidence of acetabular erosion, while Drexler et al.^[3] found that 13.8% had protrusio acetabuli and 4.8% exhibited mild degenerative changes.

The limitations of this study include its retrospective design, short-term follow-up in the THA group, and the use of different prostheses and fixation methods. Patient-reported outcomes, beyond MSTS and HHS, were not used, such as the Toronto Extremity Salvage Score or EQ-5D quality-of-life measures.

CONCLUSION

This study suggests that functional outcomes are better with the total hip compared to the bipolar proximal femoral modular prosthesis. In addition, there was no statistically significant difference between the total hip and the bipolar proximal femoral modular prosthesis in dislocation or infection rates. Precise cup positioning by arthroplasty-trained surgeons and surgeon experience may help reduce the rate of dislocations. Acetabular erosion in the bipolar proximal femoral prosthesis was higher than in previous studies, suggesting that for young patients with primary bone tumors, reconstruction with a total hip modular prosthesis might provide a better functional outcome and decrease the rate of acetabular erosion.