Numerous mechanisms may result in the loss of proximal femoral bone stock in a patient who has undergone total hip arthroplasty (THA), including infection, mechanical loosening, osteolysis secondary to particle debris, stress-shielding with adaptive bone remodeling, periprosthetic fracture (Figures 1a and 1b), and nonunion (Figures 2a and 2b). The integrity of proximal bone stock could also be compromised by insertion and removal of implants in multiple prior reconstructive procedures. Proximal femoral bone loss is often encountered during revision hip arthroplasty and can be a difficult challenge, often requiring an array of surgical techniques.

Figure 1a. Periprosthetic fracture

Figure 1b. PFR for periprosthetic fracture

Figure 2a. Nonunion

Figure 2b. PFR for nonunion

Preoperative Planning

Patient’s general health and medical management. Patients undergoing revision THA may have had multiple surgeries, as well as medical co-morbidities that should be optimized prior to undergoing revision THA. The patient’s nutritional status should be evaluated with common laboratory measurements (lymphocyte count, transferring level, and pre-albumin level); these levels should be corrected if found to be suboptimal.

Blood conservation. Because revision THA may require extensive soft tissue dissection, it is important to consider blood loss and the length of surgery when choosing the method of blood conservation. The use of a cell saver should be considered in non-infected cases. The type of anesthesia can be critical; we routinely use regional anesthesia as it provides numerous benefits compared with general anesthesia.

Current infection or history of infection. Ensuring the absence of infection prior to implantation should be the highest priority. We evaluate patients for potential infection by ordering serology: C-reactive protein and erythrocyte sedimentation rate. Patients with values suspicious for infection, such as elevated serology, or a concerning history, such as history of infections, undergo joint aspiration, with the fluid analyzed for leukocyte count, neutrophil differential, and culture.

History of previous surgical procedures and current implants. The choice of implants or technique may be altered by current implants on the femoral or acetabular side. A complete knowledge of implanted femoral devices will help determine areas of possible osseointeggration, need for extended trochanteric osteotomies, or appropriate extraction devices even in the face of extensive femoral bone loss. Many patients undergoing reconstruction have had multiple previous procedures. Examine the incision site carefully for the presence of skin lesions that may predispose the patient to infection and to determine the appropriate previous scar to be used. A new incision may occasionally need to be used if the previous scars are not accessible.

Orthopaedic examination, including assessment of motor and sensory function and vascular status.  This will be done in the office, but should be repeated the morning of surgery. Limb lengths are routinely assessed by using blocks under the shorter leg. Any discrepancy between apparent and actual limb length needs to be investigated. The reason for apparent limb equality usually is due to either fixed or flexible pelvic obliquity. If pelvic obliquity is fixed, limb length will not be restored by soft tissue release alone. This condition often requires lengthening of the operative extremity based on measured discrepancy. Patient should be counseled about the possibility of limb length discrepancy that may result from surgery. In our opinion, lengthening of the limb up to 4 cm can be carried out safely. Any lengthening beyond this point is likely to place the neurovascular structures at risk. Intraoperative monitoring of the sciatic and femoral nerves of patients may need to performed in whom extensive (>4cm) limb lengthening is anticipated.  Particular attention should be given to abductor function, which affects stability. Nonfunctional abductors could be a result of paralysis or loss of continuity and they need to be addressed at the time of femoral revision.

Radiographs. Plain radiographs may underestimate the amount of bone loss, particularly if osteolysis is present. We routinely use the Paprosky femoral bone loss classification when evaluating plain films to help assess potential reconstruction options. A crucial and often overlooked portion of the radiographic review includes an assessment of femoral deformity. Loose femoral implants are often associated with femoral remodeling into varus and retroversion. In addition, it is important to evaluate the bow of the femur on the lateral view. If there has been a femoral fracture, or if an osteotomy or another surgical procedure has been performed on the femur, anteroposterior and lateral radiographs of the entire femur are needed. As for every revision surgery, careful preoperative templating of an adequate set of radiographs is crucial. Selection of an appropriate stem length and diameter is essential to restore hip biomechanics. Despite the most accurate preoperative measurements, a variety of prosthesis size should be available, as well as multiple forms of allograft bone, in the operating room as intra-operative adjustments with change in anticipated size of prosthesis is common.


We routinely place patients in the supine position for primary and revision THA. In this supine position, leg length assessment and adjustment can be routinely and reliably performed intraoperatively. We feel that having the patient in the supine position facilitates anatomic placement of the components, as they can be easily referenced through anatomic landmarks or determined by the axis of the supine patient. A recognized limitation to operating with the patient in the supine position is that it often requires a second assistant to retract the operative limb or tissue across the patient.


We employ a modified Hardinge approach, described by Head et al, with the patient positioned supine on the operating table. Patient placement should allow maximal visualization of any previous incision, which should be utilized when possible to preserve blood supply to surrounding soft tissue. The course of the previous incision should be marked carefully with a fine point marker, and the incision should be initially centered over the greater trochanter. Dissection should be carried down to the fascia lata, cauterizing bleeding vessels along the way. Previous scar tissue will potentially make dissection and anatomic planes difficult to discern; palpation of the greater trochanter should help guide dissection.

Once the fascia has been adequately exposed, the incision should progress at anteriorly and distally, continuing in a posterior and proximal direction to cross  the natural fiber direction of the fascia. Care should be taken not to follow the anatomic course of the fascial fibers as they may lead to the posterior aspect of the femur. Once the proper plane is identified behind the fascia, the posterior fascial flap can be raised by removing any scar tissue or adhesions behind the tissue. The anterior fascial flap should be mobilized with the same precision to provide adequate exposure. Correct identification of the gluteus medius fibers running in an oblique direction should now be visible. Multiple previous surgeries may make muscle identification difficult, reinforcing the need for adequate exposure and skin incision.

A split in medius is made and a Homan retractor is placed, leaving approximately 50% of the muscle posterior. The anterior edge of the gluteus medius is identified and another Homan retractor is placed at the junction of the anterior edge of the gluteus medius and vastus lateralis. The anterior flap is raised with electric cautery, exposing pseudocapsule below while maintaining a cuff of tendon to repair along the femur. Pseudocapsule and scar can now be removed to expose the prosthesis. The hip should remain located until a thorough capsulectomy is performed, at which time hip dislocation may be advantageous.

Often the femur is scarred down and requires separate mobilization. Medial capsule and scar should be removed from the proximal femur and any bone or soft tissue overlying the shoulder of the prosthesis should be removed in order to facilitate potential prosthesis extraction.


If intact, an osteotomy to split the proximal femur may be required to facilitate the removal of the previous prosthesis and/or hardware. We prefer a longitudinal Wagner type of coronal plane osteotomy to split the proximal femur in patients with poor bone quality.  A transverse osteotomy is then made in the host bone at the most proximal area of adequate circumferential quality bone. The maximum length of native femur is maintained at all cost, because the outcome of this procedure is influenced directly by the length of the remaining femur.

Once the femur is exposed; the distal portion of the canal is prepared by successive broaching. The cancellous bone, if any, should be preserved for better cement interdigitation. After completion of femoral preparation and determination of the size of best-fit broach, trial components are inserted and the stability of the hip is examined. The best version of the femoral neck that provides stability is noted. The final femoral component is then cemented in place. A distal femoral restrictor is used whenever possible. The restrictor is introduced and advanced distally to allow for at least 2 cm of bone cement at the tip of the stem. The cement is pressurized and the final component implanted, ensuring that the porous coated portion of the stem is placed directly and firmly against the diaphyseal bone with no inter-positioning cement.

The prosthesis can be assembled and then cemented distally, or the stem can be cemented and the body then assembled onto it. In any case, extreme care needs to be exercised to prevent rotational mal-positioning.

To mark the rotation, we use a sharp osteotome to scratch the distal femoral cortex once the trial component is appropriately positioned. The rotation of the component cannot be changed once the distal stem is cemented in place. The length of the femoral component is determined through careful preoperative planning and intraoperative assessment. Two methods may be used for proper leg length determination.

  • Method one: Apply traction to the limb, with measurement from the cup to the host bone osteotomy site (for the proximal femoral replacement cases).
  • Method two (preferred): Place a Steinman pin in the iliac crest to measure a fixed point on the femur before dislocation

With the long-stem trial prosthesis in place, proper leg length can be accurately restored. For patients with total femur replacement, radiographs of the opposite and normal femur may be obtained preoperatively and used for accurate templating for length. The length of the prosthesis usually equals the length of the bone being resected, although in many of these patients, the integrity of the bone may be breached and the anatomy markedly altered. Ultimately, the femoral prosthesis length depends on the soft tissue tension about the hip. Balancing tension, restoration of the limb length, and avoiding excessive tension on the sciatic nerve is of utmost importance. 

Pearls and Pitfalls

  • Examine the patient thoroughly. Note previous scars, status of abductors, and limb length
  • Communicate with the patient and temper his or her expectations
  • Perform detailed preoperative templating. Have the company representative available to review your templating and to ensure that the correct components, and neighboring sizes, will be available on the day of surgery
  • Ensure thorough medical optimization of the patient has been carried out
  • Ask for an experienced scrub and anesthetic team
  • Minimize soft tissue dissection off the native bone and retain as much of the host bone as possible
  • Restore appropriate leg length and soft tissue tension
  • Have a low threshold for the use of constrained liners
  • Ensure good hemostasis and perform meticulous closure

Postoperative Care

Prophylactic antibiotics are given intravenously and maintained until final culture results are obtained. Thromboembolic prophylaxis is also administered. Patients are allowed to commence protective weight bearing on postoperative day one. We recommend the use of abduction orthosis for all patients and protective weight bearing for 12 weeks until adequate soft tissue healing occurs. Patients are usually able to ambulate with the use of a walking aid during this time. Patients should participate in daily physical therapy for assistance with ambulation, muscle strengthening and range of motion exercise for the knee.


In 1983, Sim et al reported on the use of megaprosthesis to reconstruct 21 proximal femurs in non-neoplastic condition. All patients experienced significant pain relief. However, there were two failures: one needed acetabular revision and the other required femoral component revision for instability.

Malkani et al retrospectively investigated the outcome of 50 revision THAs, in which a prosthetic femoral replacement was used in patients with non-neoplastic conditions. There was significant improvement in the Harris hip score at 1 year postoperatively and at the latest follow-up, as well as significant improvement in gait and ability to ambulate. In addition, 88% and 76% of patients achieved pain relief at 1 year and at the latest follow-up, respectively. However, detailed radiographic analysis revealed an increase in the number of progressive radiolucent lines around the femoral and acetabular components; aseptic loosening was the primary reason for re-revision surgery. The most common complication in this series was dislocation (22%).

Similar results were published by Haentjens et al in series of 19 hips. At 5 years, good outcome was seen in nine hips, fair outcome in five hips, and poor in two hips. The rate of dislocation was 37%, and the rate of infection was 16%. Ross et al reported a good outcome in 60% of 21 patients who underwent revision for failed total hip arthroplasty with a megaprosthesis replacement. The main complications were femoral shaft fracture in 12 hips, dislocation in nine hips, and infection in three hips.

Early experience at the authors’ institution (Thomas Jefferson University Hospital, Philadelphia, PA) in patients with failed hip prosthesis and severe bone loss for whom the only option was resection arthroplasty is encouraging. Our initial review revealed that the mode of failure of a megaprosthesis is similar in patient with or without neoplastic conditions. We were unable to detect any significant difference in the outcome of a megaprosthesis with respect to failure, incidence of radiographic lucency, limp, pain relief, and the use of walking aids between these two groups.

The senior author (JP) reported the outcomes of proximal femoral replacement for non-neoplastic conditions in 48 patients from two institutions. There was significant improvement in function, as measured with the Harris hip score. Functional outcome was found to be excellent or good in 22 patients, fair in 10 patients, and poor in 11 patients. Ten patients required a reoperation or revision because of at least one complication, which included revision for instability (closed reduction in five patients and one open reduction in one patient), revision for acetabular loosening in three cases, resection arthroplasty because of a non-reconstructable failure of constrained acetabular component in one patient, and irrigation and debridement with retention of components in one patient with periprosthetic infection. 

Klein et al, also from the authors’ institution, reported the results of proximal femoral replacement following Vancouver type B-3 fractures characterized by severe proximal bone deficiency and loose femoral stem. Of the 21 patients enrolled in this study, all but one patient were able to walk and had minimal to no pain at the time of latest follow-up (mean, 3.2 years). The rate of complications – which included persistent wound drainage that required irrigation and debridement (two hips), dislocation (two hips), re-fracture of the femur distal to stem (one hip), and acetabular cage failure (one hip) – was relatively high. Despite this, proximal femoral replacement is a viable option for the treatment of periprosthetic fractures in older patients with severe bone deficiency. If a proximal femoral replacement is used, the stability of the hip must be tested diligently intra-operatively and a constrained acetabular liner should be utilized if instability is encountered. Furthermore, the proximal part of the femur, however poor in quality, should be retained for re-approximation onto the implant to enhance the bone stock.

Al-Taki et al reported on the quality of life for patients undergoing proximal fermoral preplacements. They retrospectively reviewed 63 patients from their institution, administered various validated outcome questionnaires, and compared the results with those of a control group comprised of patients undergoing conventional revision hip arthroplasty. The authors concluded that although the outcome scores of the proximal femoral replacement patients were lower than the control group, the patient’s quality of life was significantly improved. They also found that when dislocation remains a concern, constrained liners should be used.


Early studies have reported various complications associated with the use of proximal femoral replacements:

  • Dislocation
  • Infection
  • Fracture
  • Leg length discrepancies

Dislocation, the most common major complication associated with the megaprosthesis, has reported rates ranging from 18-50%. The etiology is probably multifactorial, but most patients undergoing revision THA have had multiple surgeries, compromising the soft tissue structures around the hip. In addition many patients have multiple coomorbidities that hamper their recovery. Early prosthesis were often monolithic, making soft tissue tensioning and leg length equalization more challenging than the current second-generation modular components.

Additional reported complication include:

  • High rates of radiolucencies around the acetabulum and femur
  • Aseptic loosening
  • Catastrophic failure of the socket in patients with poor bone stock


Malkani AL, Settecerri JJ, Sim FH, Chao EY, Wallrichs. Long-term results of proximal femoral replacement for non-neoplastic disorders, J Bone Joint Surg Br 1995;77(3):351-6

Sim FH, Chao EY. Hip salvage by proximal femoral replacement. J Bone Joint Surg Am 1981;63(8):1228-39

Parvizi J, Sim FH. Proximal femoral replacements with megaprostheses. Clin Orthop Relat Res 2004(420):169-75

Parvizi J, Tarity TD, Slenker NF, Wade R. Trappler WJ, Hozack PF, Sim FH. Proximal femoral replacement in patients with non-neoplastic conditions. J Bone Joint Surg Am 2007;89(5) 1036-43

Klein GR, Parvizi J, Rapuri V, Wolf CF, Hozack WJ, Sharkey PF, Purtill JJ. Proximal femoral replacement for the treatment of periprosthetic fractures. J Bone Joint Surg Am 2005;87(8):1777-81

Al-Taki MM, Masri BA, Duncan CP, Garbuz DZ. Quality of life following proximal femoral replacement using a modular system in revision THA. Clin Orthop Relat Res 2011;469(2):470-5. 


pfr.JPG (image/jpeg)

figure1a.JPG (image/jpeg)

Figure1b.JPG (image/jpeg)

Figure 2a.jpg (image/jpeg)

Figure 2b.JPG (image/jpeg)