Previous reports on the outcomes of isolated head and liner exchange in revision total hip arthroplasty have found high rates of instability following these surgeries. Most reports have studied constructs using ≤28mm femoral heads. The purpose of this study was to determine if modern techniques with the use of larger head sizes can improve the rate of instability after head and liner exchange.

We identified 138 hips in 132 patients who underwent isolated head and liner exchange for polyethylene wear/osteolysis (57%), acute infection (27%), metallosis (13%), or other (2%). All patients underwent revision with either 32mm (23%), 36mm (62%), or 40mm (15%) diameter heads. Crosslinked polyethylene was used in all revisions. Lipped and/or offset liners were used in 104 (75%) hips. Average follow up was 3.5 (1.0–9.1) years. Statistical analyses were performed with significance set at p<0.05.

Revision-free survivorship for any cause was 94.6% and for aseptic causes was 98.2% at 5 years. 11 (8%) hips experienced a complication with 7 (5%) hips requiring additional revision surgery. Following revision, 4 (3%) hips experienced dislocation, 5 (4%) hips experienced infection, and 1 (1%) hip was revised for trunnionosis. No demographic or surgical factors significantly affected outcomes.

Our study shows that isolated head and liner exchange using large femoral heads and modern liners provides for better stability than previous reports. The most common complication was infection. We did not identify specific patient, surgical or implant factors that reduced the risk of instability or other complication.

Introduction

Re-revision due to instability and dislocation can occur in up to 1 in 4 cases following revision total hip arthroplasty (THA). Optimal placement of components during revision surgery is thus critical in avoiding re-revision. Computer-assisted navigation has been shown to improve the accuracy and precision of component placement in primary THA; however, its role in revision surgery is less well documented. The purpose of our study was to evaluate the effect of computer-assisted navigation on component placement in revision total hip arthroplasty, as compared with conventional surgery.

As the number of patients who have undergone total hip arthroplasty rises, the number of patients who require surgery for a failed total hip arthroplasty is also increasing. It is estimated that 183,000 total hip replacements were performed in the United States in the year 2000 and that 31,000 of these (17%) were revision procedures. Reconstruction of the failed femoral component in revision total hip arthroplasty can be challenging from both a technical perspective and in preoperative planning. With multiple reconstructive options available, it is helpful to have a classification system which guides the surgeon in selecting the appropriate method of reconstruction. A classification of femoral deficiency has been developed and an algorithmic approach to femoral reconstruction is presented.

An extensively coated, diaphyseal filling component reliably achieves successful fixation in the majority of revision femurs. The surgical technique is straightforward and we continue to use this type of device in the majority of our revision total hip arthroplasties. However, in the severely damaged femur (Type IIIB and Type IV), other reconstructive options may provide improved results. Based on our results, the following reconstructive algorithm is recommended for femoral reconstruction in revision total hip arthroplasty.

Type I: In a Type I femur, there is minimal loss of cancellous bone with an intact diaphysis. Cemented or cementless fixation can be utilised. If cemented fixation is selected, great care must be taken in removing the neo-cortex often encountered to allow for appropriate cement intrusion into the remaining cancellous bone.

Type II: In a Type II femur, there is extensive loss of the metaphyseal cancellous bone and thus, fixation with cement is unreliable. In this cohort of patients, successful fixation was achieved using a diaphyseal fitting, extensively porous coated implant. However, as the metaphysis is supportive, a cementless implant that achieves primary fixation in the metaphysis can be utilised.

Type IIIA: In a Type IIIA femur, the metaphysis is non-supportive and an extensively coated stem of adequate length is utilised to ensure that more than 4cm of scratch fit is obtained in the diaphysis.

Type IIIB: Based on the poor results obtained with a cylindrical, extensively porous coated implant (with 4 of 8 reconstructions failing), our present preference is a modular, cementless, tapered stem with flutes for obtaining rotational stability.

Type IV: The isthmus is completely non-supportive and the femoral canal is widened. Cementless fixation cannot be reliably used in our experience, as it is difficult to obtain adequate initial implant stability that is required for osseointegration. Reconstruction can be performed with impaction grafting if the cortical tube of the proximal femur is intact. However, this technique can be technically difficult to perform, time consuming and costly given the amount of bone graft that is often required. Although implant subsidence and peri-prosthetic fractures have been associated with this technique, it can provide an excellent solution for the difficult revision femur where cementless fixation cannot be utilised. Alternatively, an allograft-prosthesis composite can be utilised for younger patients in an attempt to reconstitute bone stock and a proximal femoral replacing endoprosthesis used for more elderly patients.

Introduction

Computer-assisted hip navigation offers the potential for more accurate placement of hip components, which is important in avoiding dislocation, impingement, and edge-loading. The purpose of this study was to determine if the use of computer-assisted hip navigation reduced the rate of dislocation in patients undergoing revision THA.

Introduction

The demands placed upon joint surgeons are perhaps greatest when treating the revision arthroplasty patient, who present with complications demanding skill in diagnosis and evaluation, interpersonal communication and the technical aspects of the revision procedure. However, little information exists identifying which specific tasks in revision arthroplasty are most difficult for surgeons to master, and whether the greatest challenges arise from clinical, cognitive or technical facets of patient treatment. This study was undertaken to identify which tasks associated with revision total knee replacement (TKR) are perceived as most challenging to young surgeons and trainees to guide future efforts in surgical training and curriculum development.

Introduction

While THA is associated with positive results and long-term improvement in patient quality of life, outcomes are nonetheless associated with adverse events and post-procedural deficits related to discrepancies in leg length (LLD), offset and cup placement. Post-THA errors in these parameters are associated with gait alteration, low back pain and patient dissatisfaction. Such discrepancies often necessitate revision and increasingly lead to medical malpractice litigation.

Maintaining accuracy in post-surgical leg length, offset and cup placement during THA is difficult and subject to error. The sensitivity of these factors is highlighted in studies that have shown that a change of as little as 5 degrees of flexion or abduction can induce alterations in leg length of up to several millimeters. Similarly, positioning of implants can alter global and femoral offset, affecting abductor strength, range of motion and overall physical function. Compounding the biochemical issues associated with inaccurate leg length are the costs associated with these deficits.

Traditional freehand techniques of managing intra-operative parameters rely on surgeon experience and tissue tensioning to manually place components accurately. These methods, however, are only able to assess leg length and are subject to inaccuracies associated with patient movement or orientation changes during surgery. Mechanical methods of minimizing post-surgical discrepancies have been developed, such as outrigger or caliper devices, although these methods also address leg length only and provide poor feedback regarding offset and center of rotation, therefore providing insufficient data to accurately achieve appropriate post-surgical leg length.

Computer-assisted navigation methods provide more data regarding leg length, offset and center of rotation, but are limited by their cumbersome nature and the large capital costs associated with the systems.

The Intellijoint HIP^® surgical smart tool (Intellijoint Surgical, Inc., Waterloo, ON) is an intra-operative guidance tool that provides surgeons with real time data on leg length, offset and center of rotation, thereby allowing for confident selection of the correct implant in order to ensure appropriate post-surgical biomechanics.

The early clinical results from an initial cohort of patients indicate that Intellijoint HIP^® is safe and effective. No adverse events were reported in the initial cohort, and the smart tool was able to measure surgical parameters to within 1mm when compared to radiographic measurements. With training cases removed, 100% of cases had a post-procedure leg length discrepancy of less than 5mm.

This paper describes the indications, procedural technique and early clinical results of the Intellijoint HIP^® smart tool, which offers a safe, accurate and easy-to-use option for hip surgeons to manage leg length, offset and cup position intra-operatively.

The unacceptable failure rate of cemented femoral revisions led to many different cementless femoral designs employing fixation in the damaged proximal femur with biological coatings limited to this area. The results of these devices were uniformly poor and were abandoned for the most part by the mid-1990's.

Fully porous coated devices employing distal fixation in the diaphysis emerged as the gold standard for revisions with several authors reporting greater than 90% success rate 8–10 years of follow-up. Surgical techniques and ease of insertion improved with the introduction of the extended trochanteric osteotomy as well as curved, long, fully porous coated stems with diameters up to 23mm. The limits of these stems were stretched to include any stem diameter in which even 1–2cm of diaphyseal contact could be achieved. When diaphyseal fixation was not possible (Type IV), the alternatives were either impaction grafting or allograft prosthetic composite (APC).

As the results of fully porous coated stems were very carefully scrutinised, it became apparent that certain types of bone loss did not yield the most satisfactory results both clinically and radiographically.

When less than 4cm of diaphyseal press fit (Type IIIB) was achieved, the mechanical failure rate (MFR) was over 25%. It also became apparent that even when there was 4–6cm diaphyseal contact (Type IIIA), and the stem diameter was 18mm or greater, post-operative pain and function scores were significantly less than those with smaller diameter stems. This was probably due to poorer quality bone.

Many of these Type IIIA and Type IIIB femurs had severe proximal torsional remodeling leading to marked distortion of anteversion. This made judging the amount of anteversion to apply to the stem at the time of insertion very difficult, leading to higher rates of dislocation. These distortions were not present in Type I and Type II femurs.

This chain of events which was a combination of minimal diaphyseal fixation, excessively stiff stems and higher dislocation rates led to the conversion to modular type stems when these conditions existed.

For the past 13 years, low modulus taper stems of the Wagner design have been used for almost all Type IIIA and Type IIIB bone defects. The taper design with fluted splines allows for fixation when there is less than 2cm of diaphysis.

The results in these femurs even with diameters of up to 26mm have led to very low MFRs and significantly less thigh pain. Independent anteversion adjustment is also a huge advantage in these modular stems. Similar success rates, albeit with less follow-up, have been noted in Type IV femurs.

Introduction

The objective of this study was to compare the performance of the Explant Acetabular Cup Removal System (Zimmer), which has been the favored system for many surgeons during hip revision surgery, and the new EZout Powered Acetabular Revision System (Stryker).

The treatment of severe acetabular bone loss is challenging, especially in the setting of an associated chronic pelvic discontinuity. There are several available treatment options for chronic pelvic discontinuity, each of which has its own disadvantages. One of the major difficulties with this entity, regardless of the reconstructive technique chosen, is the inability to obtain reproducible healing of the discontinuity. We evaluated the use of acetabular distraction, a technique which achieves peripheral or lateral distraction and central or medial compression across the discontinuity. We recommend acetabular distraction to allow for implantation of a stable construct, achieve biologic fixation and increase the likelihood of discontinuity healing.

In this multi-center trial, 32 patients that underwent acetabular revision for a chronic pelvic discontinuity using acetabular distraction were radiographically evaluated at a minimum of 25 months (range, 25 to 160 months). The study cohort was categorized according to the Paprosky acetabular bone loss classification: seven (22%) type IIC, five (16%) type IIIA, and 20 (62%) type IIIB defects. Fourteen (70%) of the 20 patients with a type IIIB acetabular bone loss pattern required use of augments for acetabular reconstruction.

Of the 32 patients, 1 (3%) patient required a revision for aseptic loosening, 2 (6%) patients had evidence of radiographic loosening but were not revised, and 3 (9%) patients had migration of the acetabular component into a more stable position. Radiographically, 22 (69%) of the cohort demonstrated healing of the discontinuity. The Kaplan-Meier construct survivorship was 83.3% when using aseptic acetabular loosening as an end-point.

During this study, the authors created a new pelvic discontinuity classification based on the type of reconstruction required. The classification mirrors the Paprosky acetabular bone loss classification. A Type I chronic pelvic discontinuity required jumbo cup reconstruction without augments. A type II discontinuity required the use of an augment for an extracavitary defect. A type III discontinuity required an augment for an intracavitary defect.

Type III defects were further subdivided into type IIIA and IIIB discontinuity. Type IIIA discontinuities utilized an augment to reconstruct the anterosuperior and/or posteroinferior column defect for primary stability of the overall construct. Type IIIB discontinuities utilized augments to reconstruct the anterosuperior and/or posteroinferior column defect for primary stability as well as a posterosuperior augment for supplemental fixation. All augments were unitized to the cup with cement. Type IV defects were massive defects that required the use of two orange-slice augments, secured together with screws and placed centrally to restore the defect, and a cup implanted and unitized to the augments with cement.

According to this new classification, the discontinuity reconstructions in our study were classified as follows: 12 (38%) type I, 8 (25%) type II, 6 (19%) type IIIA, 6 (19%) type IIIB, and 0 as type IV. Acetabular distraction technique demonstrates favorable radiographic outcomes with reproducible discontinuity healing in a majority of cases. This alternative technique allows for biologic fixation and intra-operative customization of the construct to be implanted based on the bone loss pattern present following component removal.

Bone is a dynamic organ with remarkable regenerative properties seen only otherwise in the liver. However, bone healing requires vascularity, stability, growth factors, a matrix for growth, and viable cells to obtain effective osteosynthesis. We rely on these principles not only to heal fractures, but also achieve healing of benign bone defects. Unfortunately we are regularly confronted with situations where the local environment and tissue is insufficient and we must rely on our “biologic tool box.” When the process of bone repair requires additional assistance, we often look to bone grafting to provide an osteoconductive, osteoinductive, and/or osteogenic environment to promote bone healing and repair.

The primary workhorses of bone grafting include autogenous bone, cadaver allograft, and bone graft substitutes. Among the first types of bone graft used and still used in large quantities today include autogenous and cadaver allograft bone. Allografts are useful because it is present in multiple forms that conform to the desired situation. But autogenous bone graft is considered the gold standard because it possesses all the fundamental properties to heal bone. However, it has been associated with high rates of donor site morbidity and typically requires an inpatient hospitalization following the procedure only adding to the associated costs.

The first bone graft substitute use was calcium sulfate in 1892, and over the past 122 years advancements have achieved improved material properties of calcium sulfate and helped usher in additional bioceramics for bone grafting. Today there are predominantly four types of bioceramics available, which include calcium sulfate, calcium phosphate, tricalcium phosphate, and coralline hydroxyapatite. They come in multiple forms ranging from pellets and solid blocks to injectable and moldable putty. In comparison to autogenous bone graft, the primary limitation of bioceramics are the lack of osteogenic and osteoinductive properties. Bioceramics work by creating an osteoconductive scaffold to promote osteosynthesis. The options of bone graft substitutes don't end with these four types of bioceramics. Composite bioceramics take advantage of the differing biomechanical properties of these four basis types of bioceramics to develop improved materials. To overcome the lack of osteoinductive and osteogenic properties growth factors or bone marrow aspirate can be added to the bioceramic. As a result, the list of combinations available in our “biologic tool box” continues to expand. More than 20 BMPs have been identified, but only BMP-2 and BMP-7 have FDA approval.

As we look forward to areas of future research and need within orthobiologics, some will likely come in the near future while others are much further in the future. We will continue to strive for the ideal bone graft substitute, which will have similar osteoinductive properties as autograft. The ultimate bone graft substitute will likely involve stem cells because it will allow an alternative to autogenous bone with the same osteogenic potential.

The extended proximal femoral osteotomy has been used primarily in conjunction with cementless fixation, but has been described for use with cemented stems as well. The extended proximal femoral osteotomy is indicated for the removal of well-fixed cemented and cementless implants, as well as removal of cement in patients with a loose femoral component in a well-fixed cement mantle. Although the osteotomy is not required for many femoral revisions, it is an absolute indication in patients with femoral component loosening and subsequent varus remodeling of the proximal femur.

The osteotomy diminishes the risk of an inadvertent fracture of the often compromised greater trochanter especially upon removal of a failed femoral component from its subsided or migrated position. The osteotomy enhances the exposure of the acetabulum which may be difficult in the revision setting due to multiple surgeries, severe migration of the acetabular component or the heterotopic ossification.

The extended proximal femoral osteotomy can also be used in the primary setting when a proximal femoral deformity interferes with straight reaming of the femoral canal, such as in patients with various dysplasias, previous corrective osteotomies or malunions.

As the number of patients who have undergone total hip arthroplasty rises, the number of patients who require surgery for a failed total hip arthroplasty is also increasing. It is estimated that 183,000 total hip replacements were performed in the United States in the year 2000 and that 31,000 of these (17%) were revision procedures. Reconstruction of the failed femoral component in revision total hip arthroplasty can be challenging from a technical perspective and in pre-operative planning. With multiple reconstructive options available, it is helpful to have a classification system which guides the surgeon in selecting the appropriate method of reconstruction. A classification of femoral deficiency has been developed and an algorithmic approach to femoral reconstruction is presented.

Type I:

Minimal loss of metaphyseal cancellous bone with an intact diaphysis. Often seen when conversion of a cementless femoral component without biological ingrowth surface requires revision. Type II: Extensive loss of metaphyseal cancellous bone with an intact diaphysis. Often encountered after the removal of a cemented femoral component. Type IIIA: The metaphysis is severely damaged and non-supportive with more than 4cm of intact diaphyseal bone for distal fixation. This type of defect is commonly seen after removal of grossly loose femoral components inserted with first generation cementing techniques. Type IIIB: The metaphysis is severely damaged and non-supportive with less than 4cm of diaphyseal bone available for distal fixation. This type of defect is often seen following failure of a cemented femoral component that was inserted with a cement restrictor and cementless femoral components associated with significant distal osteolysis. Type IV: Extensive meta-diaphyseal damage in conjunction with a widened femoral canal. The isthmus is non-supportive.

An extensively coated, diaphyseal filling component reliable achieves successful fixation in the majority of revision femurs. The surgical technique is straightforward and we continue to use this type of device in the majority of our revision total hip arthroplasties. However, in the severely damaged femur (Type IIIB and Type IV), other reconstructive options may provide improved results.

As the number of patients who have undergone total hip arthroplasty rises, the number of patients who require surgery for a failed total hip arthroplasty is also increasing. Reconstruction of the failed femoral component in revision total hip arthroplasty can be challenging from both a technical perspective and in pre-operative planning. With multiple reconstructive options available, it is helpful to have a classification system which guides the surgeon in selecting the appropriate method of reconstruction.

Type I: Minimal loss of metaphyseal cancellous bone with an intact diaphysis. Often seen when conversion of a cementless femoral component without biological ingrowth surface requires revision. Type II: Extensive loss of metaphyseal cancellous bone with an intact diaphysis. Often encountered after the removal of a cemented femoral component. Type IIIA: The metaphysis is severely damaged and non-supportive with more than 4 cm of intact diaphyseal bone for distal fixation. This type of defect is commonly seen after removal of grossly loose femoral components inserted with first generation cementing techniques. Type IIIB: The metaphysis is severely damaged and non-supportive with less than 4 cm of diaphyseal bone available for distal fixation. This type of defect is often seen following failure of a cemented femoral component that was inserted with a cement restrictor and cementless femoral components associated with significant distal osteolysis. Type IV: Extensive meta-diaphyseal damage in conjunction with a widened femoral canal. The isthmus is non-supportive.

Based on our results, the following reconstructive algorithm is recommended for femoral reconstruction in revision total hip arthroplasty. An extensively coated, diaphyseal filling component reliably achieves successful fixation in the majority of revision femurs and the surgical technique is straightforward. However, in the severely damaged femur (Type IIIB and Type IV), other reconstructive options may provide improved results. Type I: Cemented or cementless fixation can be utilised. If cemented fixation is selected, great care must be taken in removing the neo-cortex often encountered to allow for appropriate cement intrusion into the remaining cancellous bone. Type II: In this cohort of patients, successful fixation was achieved using a diaphyseal fitting, extensively porous coated implant. However, as the metaphysis is supportive, a cementless implant that achieves primary fixation in the metaphysis can be utilised. Type IIIA: An extensively coated stem of adequate length is utilised to ensure that more than 4 cm of scratch fit is obtained in the diaphysis. Type IIIB: Our present preference is a modular, cementless, tapered stem with flutes for obtaining rotational stability. Type IV: Cementless fixation cannot be reliably used in our experience, as it is difficult to obtain adequate initial implant stability that is required for osseointegration. Reconstruction can be performed with impaction grafting if the cortical tube of the proximal femur is intact. However, this technique can be technically difficult to perform, time consuming and costly given the amount of bone graft that is often required. Although implant subsidence and peri-prosthetic fractures (both intra-operatively and post-operatively) have been associated with this technique, it can provide an excellent solution for the difficult revision femur where cementless fixation cannot be utilised. Alternatively, an allograft-prosthesis composite can be utilised for younger patients in an attempt to reconstitute bone stock and a proximal femoral replacing endoprosthesis used for more elderly patients.

Acetabular distraction for the treatment of chronic pelvic discontinuity was first described by Sporer and Paprosky. The authors advocate the posterolateral approach for exposure of the posterior ilium and posterior column, The patient is secured in the lateral decubitus position. Following a systematic approach to surgical exposure, acetabular component removal should be performed with “cup out” osteotomes resulting in minimal iatrogenic bone loss.

Following component removal and confirmation of a chronic discontinuity determine the integrity of the remaining AS and PI columns. If porous metal augments are needed for primary stabilization, the augments are placed prior to cup insertion for reconstruction of the AS and/or PI column. Next, Kirschner (K) wires (size 2.4) are placed in the remaining AS and PI bone so that the distractor can be secured in an extra-acetabular position. The distractor is placed over the K-wires allowing for lateral or peripheral acetabular distraction and resultant medial or central compression at the discontinuity.

With the distractor in an extra-acetabular position, hemispherical reamers are used until an interference fit is achieved between the native or augmented AS and PI columns. The acetabulum should be reamed on reverse to avoid excessive removal of host bone. When the proper acetabular component size has been reached, the reamer will disengage from the reamer handle and the reamer can be used as a surrogate acetabular shell; when the acetabulum is maximally distracted, the entire construct will move as a unit.

Crushed cancellous allograft is used to bone graft the discontinuity and reamed on reverse. A revision tantalum cup is inserted with continual distraction using the distractor. Cement is applied to the augment surface prior to cup insertion in order to utilise the construct. Following cup insertion, the distractor and K-wires are removed. Adjuvant screw fixation is performed, with a minimum of 4 screws, and placing at least one of the screws inferiorly for fixation in the superior public ramus or ischium to prevent abduction failure of the construct.

In the setting of severely osteoporotic bone and inadequate screw fixation, an augment placed posterosuperiorly can be used for supplemental fixation. This augment is also unitised to the cup with cement at the same time as the liner is cemented into the cup. Bone wax is placed over the exposed tantalum surface of the posterosuperior augment to minimise soft-tissue ingrowth into the augment.

As the number of patients who have undergone total hip arthroplasty rises, the number of patients who require surgery for a failed total hip arthroplasty is also increasing. Reconstruction of the failed femoral component in revision total hip arthroplasty can be challenging from both a technical perspective and in preoperative planning. With multiple reconstructive options available, it is helpful to have a classification system which guides the surgeon in selecting the appropriate method of reconstruction. We have developed a classification of femoral deficiency and an algorithmic approach to femoral reconstruction is presented.

Type I: Minimal loss of metaphyseal cancellous bone with an intact diaphysis. Often seen when conversion of a cementless femoral component without biological ingrowth surface requires revision.

Type II: Extensive loss of metaphyseal cancellous bone with an intact diaphysis. Often encountered after the removal of a cemented femoral component.

Type IIIA: The metaphysis is severely damaged and non-supportive with more than four centimeters of intact diaphyseal bone for distal fixation. This type of defect is commonly seen after removal of grossly loose femoral components inserted with first generation cementing techniques.

Type IIIB: The metaphysis is severely damaged and non-supportive with less than four centimeters of diaphyseal bone available for distal fixation. This type of defect is often seen following failure of a cemented femoral component that was inserted with a cement restrictor and cementless femoral components associated with significant distal osteolysis.

Type IV: Extensive meta-diaphyseal damage in conjunction with a widened femoral canal. The isthmus is non-supportive.

Patella fracture after total knee arthroplasty has a variety of etiologies and has been reported to occur with an incidence ranging from 3% to 21%. Heavy patients with full flexion are at greatest risk for sustaining patella fracture. Overstuffing the patellofemoral joint with an oversized femoral component, an anteriorised femoral component or a femoral component placed in excessive extension can also overload the underlying patella. A similar phenomenon may be seen with underrsection of the patella or use of a thick button. Excessive patellar resection can predispose to patellar fracture as well. It has been demonstrated that a residual patella thickness of less than 15 mm can substantially increase anterior patellar strain. Asymmetric patellar resection can also critically alter the mechanical strength of the patella making it vulnerable to failure.

Elevation of the tibiofemoral joint line, from excessive femoral resection and hastened by posterior cruciate ligament release, will result in a relative patella baja. This can cause early patellofemoral articulation, which may result in patellar impingement on the tibial insert in late flexion and ultimately predispose the patella to fracture.

Surgical approach and soft tissue dissection should be as atraumatic to the patellar blood supply as possible to preserve the superolateral geniculate artery when performing a lateral retinacular release.

The classification used by Goldberg, et al is helpful for planning appropriate intervention:

Type I fractures: Avulsion type fractures generally involving the periphery of the patella without involving the implant.

Type II fractures: Disrupt the cement-prosthesis interfaces of the quadriceps mechanism.

Type IIIA fractures: Involve the pole of the patella with disruption of the patella ligament.

Type IV fractures: Fracture dislocations of the patella. Non-operative treatment is preferred when fractures are non-displaced.

INTRODUCTION: As the number of patients who have undergone total hip arthroplasty rises, the number of patients who require surgery for a failed total hip arthroplasty is also increasing. It is estimated that 183,000 total hip replacements were performed in the United States in the year 2000 and that 31,000 of these (17%) were revision procedures. Reconstruction of the failed femoral component in revision total hip arthroplasty can be challenging from both a technical perspective and in pre-operative planning. With multiple reconstructive options available, it is helpful to have a classification system which guides the surgeon in selecting the appropriate method of reconstruction.

DISCUSSION: An extensively coated, diaphyseal filling component reliably achieves successful fixation in the majority of revision femurs. The surgical technique is straightforward and we continue to use this type of device in the majority of our revision total hip arthroplasties. However, in the severely damaged femur (Type IIIB and Type IV), other reconstructive options may provide improved results. Based on our results, the following reconstructive algorithm is recommended for femoral reconstruction in revision total hip arthroplasty: TYPE I: In a Type I femur, there is minimal loss of cancellous bone with an intact diaphysis. Cemented or cementless fixation can be utilised. If cemented fixation is selected, great care must be taken in removing the neo-cortex often encountered to allow for appropriate cement intrusion into the remaining cancellous bone. TYPE II: In a Type II femur, there is extensive loss of the metaphyseal cancellous bone and thus fixation with cement is unreliable. In this cohort of patients, successful fixation was achieved using a diaphyseal fitting, extensively porous coated implant in 26 of 29 cases (90%) However, as the metaphysis is supportive, a cementless implant that achieves primary fixation in the metaphysis can be utilized. TYPE III A: In a Type IIIA femur, the metaphysis is non-supportive and an extensively coated stem of adequate length is utilised to ensure that more than 4 cm of scratch fit is obtained in the diaphysis. TYPE III B: Based on the poor results obtained with a cylindrical, extensively porous coated implant, our present preference is a modular, cementless, tapered stem with flutes for obtaining rotational stability. Excellent results have been reported with this type of implant and by virtue of its tapered design, excellent initial axial stability can be obtained even in femurs with a very short isthmus. Subsidence has been reported as a potential problem with this type of implant and they can be difficult to insert. However, with the addition of modularity to many systems that employ this concept of fixation, improved stability can be obtained by impacting the femoral component as far distally as needed while then building up the proximal segment to restore appropriate leg length. TYPE IV: In a Type IV femur, the isthmus is completely non-supportive and the femoral canal is widened. Cementless fixation cannot be reliably used in our experience, as it is difficult to obtain adequate initial implant stability that is required for osseointegration. Reconstruction can be performed with impaction grafting if the cortical tube of the proximal femur is intact. However, this technique can be technically difficult to perform, time consuming and costly given the amount of bone graft that is often required. Although implant subsidence and peri-prosthetic fractures have been associated with this technique, it can provide an excellent solution for the difficult revision femur where cementless fixation cannot be utilised. Alternatively, an allograft-prosthesis composite can be utilised for younger patients in an attempt to reconstitute bone stock and a proximal femoral replacing endoprosthesis used for more elderly patients.

The extended proximal femoral osteotomy has been used primarily in conjunction with cementless fixation, but has been described for use with cemented stems as well. The extended proximal femoral osteotomy is indicated for the removal of well-fixed cemented and cementless implants, as well as removal of cement in patients with a loose femoral component in a well-fixed cement mantle. Although the osteotomy is not required for many femoral revisions, it is an absolute indication in patients with femoral component loosening and subsequent varus remodeling of the proximal femur.

The osteotomy diminishes the risk of an inadvertent fracture of the often compromised greater trochanter especially upon removal of a failed femoral component from its subsided or migrated position. The osteotomy enhances the exposure of the acetabulum which may be difficult in the revision setting due to multiple surgeries, severe migration of the acetabular component or heterotopic ossification.

The extended proximal femoral osteotomy can also be used in the primary setting when a proximal femoral deformity interferes with straight reaming of the femoral canal, such as in patients with various dysplasias, previous corrective osteotomies or malunions.

Over a four year period of time, 142 consecutive hip revisions were performed with the use of an extended proximal femoral osteotomy. Twenty patients had insufficient follow-up or were followed elsewhere and were excluded from the review. The remaining 122 revisions included 83 women and 39 men. Average age at time of revision was 63.8 (26–84) years. Indications for revision were aseptic loosening (114), component failure (4), recurrent dislocation (2), femoral fracture (1) and second stage re-implantation for infection (1).

The extended proximal femoral osteotomy gave easy access to the distal bone-cement or bone prosthesis interface in all cases. It allowed neutral reaming of the femoral canal and implantation of the revision component in proper alignment. Varus remodeling of the proximal femur secondary to loosening was handled with relative ease implementing the osteotomy. Average time from the beginning of the osteotomy procedure to the complete removal of prosthesis and cement was 35 minutes. There were no non-unions of the osteotomised fragments at an average post-operative follow-up of 2.6 years with no cases of proximal migration of the greater trochanteric fragment greater than 2 mm, there was evidence of radiographic union of the osteotomy site in all cases by 3 months. Stem fixation with bone ingrowth was noted in 112 (92%) of 122 hips, stable fibrous fixation was seen in 9 (7%) and 1 stem was unstable and was subsequently revised. However, there was an incidence of 7% perforation rate of the femoral canal distal to the osteotomy site during cement removal. This was most prevalent where there was greater than 2 cm of cement plug present which was well bonded. When OSCAR was used instead of hand tools or power reamers, there were no perforations in 51 cases. There has been no failure of fixation with fully porous coated stems inserted in the canals where OSCAR had removed cement. Also, the use of OSCAR has allowed us to shorten the osteotomy, thus allowing a longer, intact isthmus to remain so that shorter stems can be used. We highly recommend the use of OSCAR in conjunction with the extended osteotomy for removal of well-fixed distal cement beyond the extended osteotomy site.

Introduction

Recurrent dislocation following total hip arthroplasty (THA) is a complex, multifactorial problem that has been shown to be the most common indication for revision THA. The purpose of this study was to classify causes of instability and evaluate outcomes based on an algorithmic approach to treatment.