Background
Surgical operations for lower limb amputation are performed in cases where necrosis of the lower limb (caused by tumors, congenital abnormalities, and vascular diseases) is observed or in cases where salvage surgery is contraindicated owing to severe trauma or infection. Most amputees are successful ambulators; hence, most are exposed to long-term prosthetic use [
1]. For ambulation of amputees using prosthesis, muscle power of the contralateral side hip joint flexor muscle and abductor muscles is more important than the power needed for the ambulation of non-amputees [
2].
The surgical approach for hip arthroplasty (HA) remains a controversial topic. Several surgical approaches are available that can be used to perform HA. The most commonly used approaches are the posterior approach (PA) and anterolateral approach (ALA), each with distinct advantages and disadvantages. The PA is considered easier to perform and is generally a quicker procedure, limiting operative complications such as blood loss and anesthetic issues. The abductor muscles are not disturbed significantly, so there is generally no gait abnormality [
3]. Nevertheless, PA has several limitations. The main drawback is the damage to the short external rotators of the hip, which increases the risk of postoperative instability [
4]. Acetabular exposure is limited, and it is known that there is a risk of sciatic nerve damage [
5] especially when extensive hip revisionary surgery is performed with PA [
6].
An advantage of ALA is a decreased incidence of dislocations. However, the ALA has some drawbacks. During the approach, the anterior part of the gluteus medius muscle makes it difficult to visualize the acetabulum and femur sufficiently. Therefore, tenotomy is essential [
7] and it is possible to attenuate abduction strength after surgery. Moreover, in this approach, there is a risk of injury to the inferior branch of the superior gluteal nerve [
8]. If this nerve is damaged, abduction weakening becomes more severe [
9], and the patient may be forced to limp walking after surgery [
10]. Therefore, this may severely decrease the patient’s satisfaction with the procedure [
11]. Efforts have been made to minimize damage to the gluteus medius muscle. Bertin [
10] and Higuchi et al. [
12] introduced a minimally invasive ALA that approaches the plane between the gluteus medius muscle and tensor fascia lata that minimizes gluteus medius damage. Minimally invasive ALA provides rapid rehabilitation after surgery and also has the advantage of being able to prevent posterior dislocation due to the lack of damage to the posterior articular capsule. However, it is not widely used because it has a steep learning curve as well as problems due to the high soft tissue tension during the procedure. Fortunately, muscle weakness due to damage of the gluteus medius muscle from using ALA is known to be temporary. Winter et al. [
13] performed total hip arthroplasty using PA, ALA, and the anterior approach and reported measurements of muscle strength. The leg press power and abduction strength were significantly lower 6 weeks postoperatively in cases using ALA rather than PA, but there was no difference after 3 months. The function of the gluteus medius is more important when considering the gait characteristics of the limb amputee, which support mainly the single limb. Therefore, when performing HA in the contralateral hip of the amputated lower limb, it is important to consider the injury of the gluteus medius muscle, which is known as the powerful abductor muscle in the surgical approach. This is because, as was observed in Winter’s study, if muscle strength is reduced for a certain period after surgery, it is likely that a lower limb amputee at high-risk of falling is likely to experience a fall that can have a significant impact on postoperative quality of life during this period. Herein, we present a retrospective series of HAs after contralateral lower extremity amputation. Although the authors’ study did not directly measure muscle strength postoperatively in each period, it is believed that each approach (PA and ALA) possibly affects the postoperative hip joint function, recovery of gait ability, and occurrence of accidents such as falls.
Methods
The study included patients with leg amputations who underwent total hip arthroplasty for the contralateral hip joint from January 1999 to November 2014, under four hip arthroplasty specialists affiliated to four different hospitals. The patients who had undergone HA less than 5 years previously and patients who could not walk independently and were unable to perform all social activities prior to surgery were excluded. Operations via the PA were performed with the patient in the lateral decubitus position. The short external rotator muscles, including the piriformis tendon and the capsule, were resected with a single flap. The joint was reduced after the cups and stems were inserted. The range of motion, torsion, and stability of the soft tissues were examined, and the tissue flap containing the articular capsule was directly sutured to the posterior part of the proximal femur. In this study, all ALAs were performed with the patient under the lateral decubitus position. After the subfascial space is entered, the gluteus maximus is split via blunt dissection, and a smaller dissection of only one-half to two-thirds of the gluteus medius and gluteus minimus from the anterior border of their insertion to the greater trochanter is created. The hospitals where the authors perform surgery using a computerized common medical record system and apply the same postoperative management manuals for the same operations. The patient data were collected and analyzed retrospectively using the computerized medical records of each hospital.
Perioperative evaluation
Perioperative blood loss volume (intraoperative blood loss volume was calculated by subtracting the normal saline volume used for irrigation at the total fluids volume contained in suction drain bottles and adding weight gain of gauzes used at the time of surgery; intraoperative blood loss volume was calculated by adding postoperative drain volume), operation time, and the postoperative blood transfusion volume were examined. The total hospital periods from admission to discharge were compared.
Functional evaluation
After the surgery, the timing of the start of ambulation using walking aids such as walkers or crutches was evaluated. The results of functional recovery were evaluated immediately after surgery, 3 months, 6 months, 1 year after surgery, and then annually afterward. The Harris Hip Score (HHS) [
14] for pain, ambulation, and degrees of movements and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [
15] were used for the evaluation. Postoperative activities of daily living (ADL) scale [
16] were used for functional recovery check.
Complications
To compare early stage ambulation stability after surgery, the patients who experienced a fall from surgery to 1 year were evaluated. A fall is defined as “an event which resulted in the person coming to rest inadvertently on the ground or other level, other than as a consequence of lost consciousness, a violent blow, stroke or epileptic seizure” [
17]. Moreover, infections, dislocation, periprosthetic fractures (PPFs), implants loosening, and revision surgeries that may or may not affect ambulation were evaluated. Also, we compared the results of hip arthroplasty in the amputee group with the nonamputee group with regards to the incidence of falls, dislocation, PPFs, and loosening. The incidence of complications between the two groups was compared and analyzed by matching them based on variables such as the surgeon, duration of the operation, age, approach, and inserted implants.
Statistical analysis
Statistical analyses were conducted with IBM-SPSS 18.0 software (IBM-SPSS, Armonk, NY, USA). For comparison between groups, we used a non-parametric Mann–Whitney U-test for continuous variables (such as preoperative evaluation and functional evaluation) and the chi-square test for categorical variables (such as fall, dislocation, and PPFs). Kaplan-Meier survival analyses were conducted using dislocation, periprosthetic fractures, and implant loosening as the end point for comparisons within each group. Statistical significance was set at p < 0.05.
Discussion
There is a high risk of arthritis in the contralateral hip joint in patients with below knee amputations [
19‐
21]. Kulkarni [
20] et al. analyzed 44 cases of below knee amputations and reported that 18% of patients developed arthritis in the contralateral hip joint; this was double the incidence observed in non-amputated patients. Struyf et al. reported that hip arthritis on the contralateral side of the amputated leg was five- to ten-times higher and the progression of the arthritis was faster in patients with amputations than in those without [
21]. The most common cause of arthritis in the contralateral hip joint in patients with below knee amputation was a change in their gait pattern owing to special circumstances, such as leg amputation and use of prosthetics (increased metabolic energy expenditure, decreased walking speed, larger stride width, shorter stride length with the intact limb, and increased stance time). It is known to induce pain and degenerative changes by inducing a higher load on the joint owing to an increase in the ground reaction forces. [
2,
20,
22‐
24] The gluteus medius muscle is an important abductor muscle [
25] that plays important roles in stabilizing the pelvis during the period of single support of the gait cycle [
22], and in the balance and normal movement of the pelvis and lower limb during gait [
26]. The role of the muscle abduction in the hip contralateral to the amputated leg becomes particularly more important for stable gait in patients with leg amputations. In this case, the anterior approach was used and the anterior part of the gluteus medius—which plays a major role in abduction, internal rotation, and flexion of the hip joint—was cut off [
27]. Patients who underwent the anterior approach showed more falls within the postoperative 3 months, and slower functional recovery compared to those who underwent the posterior approach. Therefore, it was thought that damage to the anterior part of the gluteus medius by the anterolateral approach may have affected the outcomes.
The incidence of periprosthetic femoral fractures after hip arthroplasty is increasing [
28]. A recent study showed that the incidence of periprosthetic femoral fracture is approximately 1% after primary HA [
29]. The authors observed a higher incidence of periprosthetic femoral fractures in both groups (3% in the PA group (1 in 33 patients) and 8.8% in the ALA group (3 in 34 patients)). It is a well-known fact that a patient with amputation belongs to the high-risk group for fall [
30]. Falls occurred more frequently in the ALA group (2 out of 3 patients) than in the PA group within 3 months after the operation. We believe that the delayed functional recovery of the ALA group (until 3 months postoperative) seems to be related to these results and the damage of the anterior fiber of the gluteus medius during ALA.
In a study of functional recovery based on the approach type, Jeya et al. [
31], in a medium term (5 years), found no difference in the clinical benefit of surgery as defined by the change in Oxford Hip Score or in the absolute postoperative Oxford Hip Score between patients who underwent PA and with those who underwent ALA. However, the initial difference at 1 year in Oxford Hip Score between the PA and ALA groups may be attributed to the increased trochanteric pain [
32] and increased gait abnormalities [
33] in the ALA group during the immediate postoperative period. In particular, Pfirrmann et al. [
32] found changes in the abductor muscle after hip arthroplasty using MRI. In the case of HA with partial incision of the gluteus medius and gluteus minimus, defects in the gluteus minimus and gluteus medius were observed in 8 and 16% of patients without postoperative trochanteric pain or limp symptoms, respectively; however, gluteus medius and gluteus minimus defects were found in 62% and in 56% of patients with symptoms such as trochanteric pain or limp, respectively. In the present study, functional recovery was lower in the ALA group than in the PA group until 3 months postoperatively, but no significant difference was observed between the two approaches from 6 months to the last follow-up. Therefore, we concluded that when HA is performed to the contralateral side hip joint of amputees, minimizing the gluteus medius damage and doing the best to repair it was necessary when using the ALA.
Amputees are a high-risk group for fall. Kulkarni et al. found that 60% of amputees reported that falling affected their daily life, work, leisure, and confidence [
30]. These falls are due to altered lower limb mechanics; therefore, transtibial amputees make compensatory gait adjustments. In the present study, gluteus medius muscle damage in patients with high-risk falls could be an important risk factor for falls and PPF in the ALA group within 3 months after surgery.
Other notable findings are those of four patients with PPF (around the femoral stem) and of two patients with dislocation due to prosthetic leg-related falls, despite the absence of osteoporosis or problems with walking ability. Therefore, providing thorough education to patients to wear and use the prosthesis with caution after surgery is important. Since ALA and PA both have advantages and disadvantages, we do not believe that only one approach should be used exclusively for hip arthroplasty in the contralateral hip joint of below the knee amputees. However, the following points should be considered before surgery. First, it is important to use the approach that is the surgeon is most familiar with. Surgeons using ALA should minimize the damage of the gluteus medius muscle and the muscles around the hip joint and should operate quickly and safely. Surgeons using PA should reduce damage and do their best to repair structures that could affect the stability of the hip joint, such as the short external rotator. Second, as our results have shown, the risk of falls should be adequately explained to patients as there is a high risk of falls and fractures around the femoral stem in amputees compared with non-amputees. In particular, patients with ALA should be more careful because if they fall within the first 3 months after surgery they will have a slow recovery of gait ability. Last, patients with a high risk of falls (those living on the floor, those not expected to be well coordinated with postoperative care, those who need to return to active work soon after surgery, etc.) should have PA performed by a skilled surgeon.
This study has some limitations. First, the sample size is small. However, collecting data from many cases is difficult because hip arthroplasty of the contralateral side hip joint in patients with below knee amputation is rare. Second, the surgery was performed by 4 surgeons all of whom have > 10 years of experience with total hip arthroplasty and have performed > 300 surgeries per year. Third, the difference in muscle strength due to the injury to the gluteus medius, which was the most significant difference between the two approaches, was not identified. It was evaluated by comparing only events of falling or functional recovery. The study was further limited by its retrospective nature and the relatively small number of patients. Therefore, the results need to be supplemented by large-scale prospective studies. The final limitation is that the study included both patients with total hip replacement and with hemiarthroplasty. In particular, 9 out of 10 hemiarthroplasty patients had hemiarthroplasty due to femoral neck fracture. Most patients with femoral neck fracture were elderly and likely to have osteoporosis and often had reduced gait ability before fracture. Yet, because we selected patients who were socially active with prosthetics before the fracture, they were included in the study. It is, however, considered that a limitation of this study was the failure to distinguish the presumed complications that are more likely to occur in total replacement, such as dislocation.