Total joint replacement is now considered a routine surgical option for small animals with advanced joint disease. This review highlights the current state of the art in total hip, elbow and knee replacement in small animals, as well as the potential for application to other joints such as the intervertebral discs, hock and shoulder. Advances in cementless fixation, the use of less-invasive, bone-sparing procedures, and the development of custom implant options for revision surgery and oncologic reconstruction will be discussed. As growth in this field continues, it is important that future developments in total joint replacement are built on a foundation of detailed laboratory evaluation of new implant designs and surgical techniques, objective assessment of clinical outcomes, and a comprehensive, coordinated retrieval program that evaluates both well-functioning and failed joint replacements.
Total joint replacement (TJR) is now widely accepted as a standard treatment for advanced degenerative joint conditions in humans. As in much of veterinary orthopaedics, many of the developments in TJR in small animals have been made possible through information gleaned from experience in human TJR. Before discussing the various options for TJR in small animals, it is appropriate and beneficial to review the history and evolution of this groundbreaking technology in humans.
TJR in humans
The first hip and knee replacements were performed in the late 19th century by the German surgical pioneer, Themistocles Gluck, who implanted ivory ball-and-socket hip replacements or hinged knee replacements in patients with tuberculous arthritis (Eynon-Lewis and others 1992). However, it is generally accepted that the era of modern TJR started with the work of the Judet brothers in France, Austin Moore in the United States and John Charnley in the United Kingdom, who was the first to report on the use of acrylic bone cement to secure a TJR (Charnley 1961, Gomez and Morcuende 2005). Charnley's initial experience with a “low-friction” arthroplasty was not encouraging, with a high incidence of implant failure secondary to aseptic loosening of the implant (Charnley 1961, Kurtz 2004). However, replacement of the original polytetrafluoroethylene (Teflon®) acetabular cup with a cup made from high-density polyethylene (HDPE) resulted in a reduction in implant wear and a -concomitant improvement in clinical outcomes (Charnley 1972). Although there has been some evolution in implant design and material selection, in particular the transition from HDPE to ultra-high-molecular-weight polyethylene, UHMWPE, the standard cemented Charnley hip replacement remains a benchmark against which all other joint replacements are compared. In the five decades since Charnley's first report, total hip replacement (THR) has evolved into one of the most successful surgical procedures in human medicine in terms of improvements in quality of life (Jones and others 2000, Ethgen and others 2004). When used in patients older than 65 years of age, 15-year survival rates for contemporary cemented THR implant designs are reported to be approximately 85% (Mäkelä and others 2008). In recent years there has been a dramatic push towards cementless fixation (Dutton and Rubash 2008, Mäkelä and others 2008) and more-conservative, less-invasive THR surgeries that involve the use of either smaller incisions or modified, muscle-sparing surgical approaches. In addition, bone-conserving procedures such as hip resurfacing have gained popularity amongst younger, more active patients who are looking to hip resurfacing as an alternative to traditional THR (McMinn and others 2011).
The history of total knee replacement (TKR) follows a similar pathway of development, with the early work having been performed by Shiers in the 1950s (Shiers 1954) and by McKee and Walldius in the 1960s (Wilson and others 1980). Clinical experience with these fixed-hinge implants revealed an unacceptable rate of implant failure (hinge breakage) or implant loosening because of direct load transfer through the hinge mechanism to the cement-bone interface (Jones and others 1979). Attention subsequently focussed on the development of improved prostheses that better restored the complex kinematics of the knee, including internal/external rotation and roll-back. This ultimately led to the development of unlinked, “condylar-style” knee implants. The archetypal condylar knee, the “Total Condylar” implant developed by John Insall in New York, consisted of a convex metallic femoral component with a geometry that was based on that of the natural distal femur, a concave UHMWPE tibial insert, and a metal tibial baseplate that secured the insert on the proximal tibia. Success rates with this implant approach 95% at 15 years (Ranawat and others 1993) and although there have been incremental advances in implant design since then, the original condylar knee remains the foundation of current TKR implant designs. Success rates of primary TKR have not changed much, but there has been a dramatic increase in the availability of modular and revision implants. As with THR, interest in minimally invasive procedures has led to the development of true minimally invasive knee implants that can be used with mini-incisions (Bonutti and others 2010). Unicompartmental knee replacement solutions that focus on replacing only the diseased joint compartment are also available and offer a viable first-line alternative to TKR (Pandit and others 2011).
Success in total elbow replacement (TER) has proved to be more elusive. As with the knee, early designs focussed on the development of a hinged implant that could be secured to bone with cement. Similar to the knee, clinical results with these fixed-hinge implants were not encouraging, with failures typically being the result of fatigue failure at the hinge mechanism (Seitz and others 2010) or aseptic loosening due to failure of the cement-bone interface (Bryan 1977). Refinements in implant design, particularly with regard to the ability to accurately recreate the rotational movements of the human elbow, have led to improvements in clinical success, although it should be noted that 5- and 10-year revision rates for TER are typically higher than for either THR or TKR (Fevang and others 2009).
Indications for TJR
In humans, the most common indications for TJR are pain and loss of function associated with primary or secondary osteoarthritis and rheumatoid arthritis (Crawford and Murray 1997). In small animals, the most common indication for TJR is degenerative joint disease (DJD), usually secondary to hip or elbow dysplasia, rupture of the cranial cruciate ligament or other intra-articular pathologies such as osteochondrosis or chronic fractures. As surgeons have become more confident with the procedure, indications have expanded, and animals with severe but not end-stage DJD are now being treated with TJR. However, there have not been any randomised studies comparing clinical outcomes in dogs with DJD that are managed by TJR versus non-surgical management or more conventional reconstructive procedures. This is critically important information, and prospective, randomised studies are needed to determine whether there are any advantages (or disadvantages) to the selection of TJR as an alternative to conservative management or a reconstructive procedure in these dogs.
At the current time, surgery is considered appropriate for dogs with moderate to severe DJD or for those with evidence of significant, full-thickness cartilage loss with bone-on-bone contact. Animals should have failed conservative management with exercise and weight control, non-steroidal anti-inflammatory drugs and physical rehabilitation. A history of prior joint surgery is not considered a contraindication to surgery, although any animal with implants in/around the joint (e.g. toggle pin for hip luxation, or a lateral suture or tibial plateau leveling osteotomy (TPLO) plate adjacent to the stifle) should be evaluated carefully to rule out the presence of infection, which is an absolute contraindication for TJR. It is current practice of the author to perform synoviocentesis on any dog with a history of prior joint surgery; if implants are present, it is recommended that the implants be removed and submitted for microbial testing before admitting the dog for TJR.
Current status of TJR in small animals
Total hip replacement
Detailed reviews of the history of canine THR have been published previously (Olmstead 1987, 1995a, 1995b) and need not be reprised here. A brief summary is, however, pertinent because the history of canine THR provides the appropriate backdrop for discussions about the development of new joint replacement solutions, many of which have not been reviewed previously.
Although a number of prototype canine THR implants were evaluated in the early 1970s (Lewis and Jones 1980), the first report of a successful clinical THR procedure was published in 1974 (Hoefle 1974). The implant used in that original report was a fixed monobloc cobalt-chromium (CoCr) alloy femoral component paired with a high-molecular-weight polyethylene acetabular component (Richards Canine II; Richards Medical, Memphis, TN) (Fig 1A). Clinical results with the Richards implant were generally very good, with success rates ranging from 85 to 95% (Leighton 1979, Lewis and Jones 1980, Olmstead and others 1983, Parker and others 1984, Olmstead 1987, Paul and Bargar 1987) but the surgical instrumentation was considered suboptimal and the use of a fixed head component limited the surgeon's ability to address intraoperative laxity by adjusting the head-neck offset (Olmstead 1995b). To address this, a modular system implant was developed and introduced by BioMedtrix in 1990. The original modular hip system, the Canine Modular Hip (BioMedtrix, Boonton, NJ, USA), comprised a titanium alloy, collared femoral component, a CoCr femoral head (17 mm in diameter) and a UHMWPE acetabular cup designed for cemented fixation. In 1994, the titanium femoral component was replaced with a CoCr component of the same design. This implant system is now known as the CFX™ (Cemented Fixation) total hip system (Fig 1B).
Additional cemented THR implants have been developed, including the Biomécanique and its derivative, Porte, and remain in use in select centres around the world. In a retrospective review of 277 dogs implanted with the cemented titanium Biomécanique implant, clinical recovery was judged to be good to excellent in 90·3% of cases. Complications were seen in 22% of dogs, with cup loosening and cup wear being the most common causes for revision surgery (Bardet 2004). In a second, larger clinical study involving 686 implants, good to excellent results were seen in over 97% of cases (Matis and Holz 2004).
The most significant recent development in canine THR has been the introduction of cementless fixation. The first cementless THR design to be used in dogs was a research version of the human porous-coated anatomic (PCA) hip, manufactured by Howmedica (Hedley and others 1987). Although never commercialised for the veterinary market, the Canine PCA hip system was used in a number of preclinical research studies (Schiller and others 1993) and was subsequently made available to a limited number of veterinary referral sites for use in clinical cases (Fig 1C). The encouraging results seen with cementless THR in dogs (Marcellin-Little and others 1999) subsequently drove the development of a commercial cementless THR system, the BFX™ (Biologic Fixation) total hip (BioMedtrix, Boonton, NJ, USA), that consists of a collarless CoCr femoral component and a hemispherical CoCr acetabular cup with an inner liner made of UHMWPE. Porous ingrowth is achieved through the use of CoCr beads (mean diameter 300 µm) over the entire backside of the acetabular component and on the shoulder of the femoral component (Fig 1D). Excellent clinical results have been reported in a large clinical series of BFX hips (Liska 2004a) and objective assessment of limb function has confirmed the efficacy of the implant in normalising hind limb loading after surgery (Lascelles and others 2010).
An alternative approach to cementless femoral fixation was developed by Tepic and Montavon as the Zurich cementless THR (Tepic 2004) and is now marketed as the Kyon hip (Kyon AG, Zurich, Switzerland) (Fig 2). The design philosophy behind the Kyon hip involves the use of transcortical locking screws to provide immediate stability after surgery, and a low-modulus, plasma-coated titanium acetabular shell to facilitate bone ingrowth. A large multi-centre clinical trial of the original (first generation) Kyon implant system identified a significant risk of postoperative complications, with the most common complications being craniodorsal coxofemoral luxation (approximately 10% of cases), femoral fracture (5%), stem failure (5%) and cup loosening (4%) (Boudrieau 2004). -Substantial modifications were made to the implants, including conversion from a titanium alloy acetabular shell to a titanium component with radial ridges to enhance press-fit, addition of a plasma coating to promote osseous integration, and changes in the design of the femoral component. The surgical technique and instrumentation were also refined by the addition of acetabular reamers with reduced manufacturing tolerances and introduction of an updated positioning guide to ensure a more accurate acetabular version. In a retrospective study on 65 second-generation implants followed for a mean of 23 months, the overall complication rate was reported to be 17%. All but two of these cases were revised successfully and the final outcome was considered successful in 97% of dogs (Guerrero and Montavon 2009). More recently, Vezzoni reported an overall 91% success rate in a series of 1053 Kyon hips performed between 2001 and 2011 (Vezzoni 2011).
The Helica Hip (Medicatech VET, Irvine, CA, USA) was introduced in 2006 as a bone-preserving THR option. The design of the Helica is based on a threaded titanium alloy femoral implant that screws into the femoral neck after resection of the femoral head (Fig 3). A stainless-steel femoral head (18 mm in diameter), coated with titanium nitride to reduce friction and wear, articulates with a UHMWPE acetabular liner enclosed in a titanium alloy acetabular shell. Fixation of the titanium alloy acetabular component is achieved through a combination of immediate -stability, achieved through self-cutting threads on the backside of the component, and osseointegration through bone attachment to the roughened surface of the acetabular component. -Short-term clinical results from the Helica have been reported as being encouraging (Hach and Delfs 2009) although objective outcome data are currently lacking.
The most recent development in THR options has been conservative hip replacement through hip resurfacing. Hip resurfacing has had a long and rather mixed history in human medicine, although it has seen something of a resurgence in recent years, largely as a result in improvements in manufacturing processes for metal-on-metal articulations (Mont and others 2006). A -metal-on-metal canine hip resurfacing implant (Bionic Hip Resurfacing System; Securos, Fiskdale, MA, USA) (Fig 4) has now been -developed and is currently undergoing clinical trials in select veterinary referral centres. The potential advantages of hip resurfacing include bone conservation (allowing for potential conversion to THR at a later date), shorter operative times, faster recovery, more normal physiologic loading of the proximal femur and a reduction in the incidence of dislocation and femoral fractures. However, hip resurfacing can be hard to perform in cases with severely dysplastic hips or in dogs with marked femoral sclerosis. There have not been any clinical studies on hip resurfacing in dogs and it will be interesting to see whether the potential advantages of this procedure are in fact realised in dogs. It is also important to recognise that the use of metal-on-metal bearings in human THR remains controversial in human orthopaedics because of concerns over potential metal sensitivity and inflammatory osteolysis (Fary and others 2011).
There have now been two published reports on the use of THR as an alternative to femoral head and neck ostectomy in cats (Liska 2010, Witte and others 2010). All of the cases to date have made use of a cemented micro-THR implant system (BioMedtrix, Boonton, NJ, USA).
Total elbow replacement
Although there have been isolated reports on the use of constrained (fixed-hinge) elbow implants in a cat (Whittick and others 1964) and in dogs (Lewis 1996), the first commercial unlinked, semi-constrained TER implant was designed by Conzemius and marketed as the Iowa State elbow (BioMedtrix; Boonton, NJ, USA). This implant comprised a CoCr humeral component that articulated with a monobloc UHMWPE radioulnar component (Fig 5A). Surgical implantation of the Iowa State elbow involves exposure of the joint via a lateral arthrotomy and implantation of both humeral and radioulnar components with polymethylmethacrylate (PMMA) bone cement. Initial experience in 6 research dogs (Conzemius and others 2001) led to a clinical trial in which satisfactory clinical results were reported in 16 of 20 dogs that underwent TER. Statistically significant improvements were reported in peak vertical force and vertical impulse in the operated limb (Conzemius and others 2003). Serious complications were seen in four dogs, including infection (two dogs), lateral luxation (one dog) and humeral fracture (one dog). A similar pattern of complications was also reported from a clinical trial in the United Kingdom (Innes 2009). In response to concerns over implant stability and the potential for periprosthetic fracture, the implant design and instrumentation underwent significant modifications. The humeral component was redesigned with a reduced shoulder to allow for preservation of bone stock and with distal porous coating to allow for a combination of cement fixation around the stem and cementless fixation around the shoulder. The surgical instrumentation was also refined to allow for more precise identification of the centre of rotation of the joint and to improve joint stability by more precise matching of the humeral and radioulnar geometries (Conzemius 2009). The second-generation Iowa elbow is currently undergoing clinical evaluation.
A cementless, semi-constrained TER system (TATE Elbow; BioMedtrix, Boonton, NJ, USA) has been developed by Acker and van der Meulen (Acker 2009). Like the Iowa State elbow, the TATE is a two-component system consisting of a CoCr humeral component and a metal-backed UHMWPE radioulnar component. In contrast with the Iowa State elbow, the TATE is designed for cementless fixation through the use of porous CoCr beads (as used on the femoral components of the BFX hip and the Canine TKR systems) applied to the backside of the humeral and radioulnar components. Implantation of the TATE is performed via a medial epicondylar osteotomy, leaving the origin of the medial collateral ligament entire. The TATE makes use of a milling system that allows for simultaneous preparation of both humeral and radioulnar bone beds. An alignment jig is used to maintain the elbow in a fixed position while the milling is performed; in this way, congruency between the two articulations is ensured. A patented “cartridge implant”, consisting of the two components linked together with a temporary spacer (Fig 5B), is then placed into the milled cavity and implanted using press-fit (Fig 5C). Clinical results with the TATE elbow have been encouraging (Acker 2009) but much remains to be learned, particularly regarding the restoration of joint kinematics. In a recent preclinical study, dynamic radiostereometric analysis was used to quantify normal elbow joint kinematics (Guillou and others 2011) and a prospective study comparing kinematics before and after TATE elbow replacement is currently underway.
A new semi-constrained implant, the Sirius elbow, has been developed at the University of Liverpool and is now manufactured by Osteogen Ltd. The implant consists of a cemented CoCr alloy humeral component and a cementless titanium/UHMWPE radioulnar component that is secured to bone with screws (Innes and others 2012) (Fig 5D). The implant has been used in three clinical cases to date and short-term results are reported to be encouraging.
A number of unicompartmental solutions to TER have been designed and then abandoned, but more recently Schulz and others (2011) published preliminary results with a Canine Unicompartmental Elbow (CUE) arthroplasty system that is designed to address medial compartment disease in dogs. The implant, consisting of a metallic humeral component and a UHMWPE ulnar plug, is implanted via a medial approach with either tenotomy or medial epicondylar osteotomy (Schulz and others 2011). Preliminary results from a multi-centre study of 22 dogs with follow-up of greater than six months indicate good short-term function, with 17 of 22 dogs (77%) demonstrating good or acceptable elbow function as determined by subjective assessment of lameness. Force plate and pressure walkway analyses are ongoing but complete data are not yet available.
Total knee replacement
Early experience with TKR in the veterinary field came through preclinical research studies performed in dogs and reported in the human orthopaedic research literature (Bobyn and others 1982, Turner and others 1989, Berzins and others 1994, Sumner and others 1994). These early studies focussed on techniques for cementless implant fixation, since this was an area of significant interest and controversy in humans.
The first commercial TKR implant designed and manufactured specifically for dogs was implanted in 2005 in a working dog that had sustained a chronic, comminuted femoral non-union secondary to a gunshot injury (Liska and others 2007). Although the bone defect in that dog necessitated the use of a custom femoral implant, the geometries of the femoral and tibial component were based on a design that was subsequently released as a commercial implant (Canine Total Knee; BioMedtrix, Boonton, NJ, USA). This implant, which has now been used in almost 300 cases worldwide, consists of a CoCr femoral component with a stem, and a monobloc UHMWPE tibial component with a keel to resist rotational and translational forces. The fixation surface of the femoral component is covered with sintered CoCr beads, allowing for cementless or cemented fixation. The UHMWPE component is designed for use with PMMA bone cement (Figs 6A,B).
The surgical procedure for canine TKR has been described previously (Allen and others 2009, Liska and Doyle 2009). As in any surgical procedure, there are a number of important technical elements in successful TKR surgery. The most important step in TKR is accurate preparation of the bone surfaces so that the femoral and tibial implants are aligned parallel to one another in the frontal plane. Failure to restore parallelism between the articulating surfaces results in joint laxity, instability and asymmetric loading of the implant. Human clinical studies have demonstrated that incorrect alignment of the implants increases the risk of implant failure (Ritter and others 1994).
A second critical step in TKR relates to preservation of the soft tissues and maintenance of appropriate soft tissue -tensioning around the joint. The medial and lateral collateral ligaments play a critical role in ensuring joint stability after TKR. Although both collaterals are at risk of injury during the preparation of bone cuts in TKR, the medial collateral is more often damaged because it is harder to visualise amongst the thickened periarticular soft tissues that constitute the medial buttress. The risk of ligament injury can be reduced by paying close attention to both identification and protection of the ligaments during the osteotomies.
Recently, an all-cementless implant system has been developed (GenuSys Knee; Innovet, Hamburg, Germany). This system uses a CoCr femoral component and a CoCr tibial baseplate that holds a UHMWPE insert. The articulating surfaces of the metal components are coated with a wear-retardant coating of titanium nitride, while the fixation surfaces are coated with plasma-sprayed titanium. The femoral component is secured to the distal femur with a bone screw and the tibial component is implanted in press-fit fashion (Figs 6C,D). Both implant systems (GenuSys and Canine Total Knee) incorporate modular implant sizing to permit use in animals with a broad range of body shapes and sizes.
There has been a single report on TKR in a cat (Fitzpatrick and Asher 2010). The implant in this case was a custom-constrained, fixed-hinge knee. Other cases are anticipated but long-term follow-up is not currently available.
Recovery following TKR is typically slower than is seen after THR, with most animals taking around 10 to 12 weeks to return to optimal joint function. Guidance has been published regarding the use of rehabilitation therapy in the postoperative care of canine TKR patients (Liska and Doyle 2009). It is hard to come up with prescriptive guidelines that work for every dog; in general, cage rest is recommended by the author for the first two weeks followed by gradual introduction of controlled, on-leash walking. Animals are assessed at 6 weeks and 12 weeks postsurgery. If the knee is stable, pain-free and able to move through a satisfactory range of motion (approximately 100°), the owners are allowed to reintroduce off-leash exercise. Dogs are examined again at 6 and 12 months, then annually thereafter.
Anecdotal feedback on the clinical results with the hybrid Canine Total Knee implant has been generally encouraging, but to date there has been only a single published report of clinical results following canine TKR in dogs with end-stage stifle DJD (Liska and Doyle 2009). In that series of six dogs, limb function (as assessed by range of motion and force plate analysis) was significantly improved one year following surgery. Additional, larger case series are now needed to more completely define clinical outcomes and the incidence and nature of postoperative complications. A summary of clinical outcomes from a series of 40 dogs was presented at the recent American College of Veterinary Surgeons meeting (Preston and Lidbetter 2011), so data to address this issue should be forthcoming over the next few months.
A cementless, metal-backed version of the tibial component from the Canine Total Knee system has been evaluated in a limited series of purpose-bred research dogs (Allen and others 2010) and this implant is now available as a custom order for clinical cases. A prospective, randomised clinical trial of cemented versus cementless fixation is currently ongoing at The Ohio State University.
Clinical follow-up on the GenuSys knee was not available in time for inclusion in this review, although a manuscript is reported to be in development (Manssur Arbabian, INNOPLANT Medizintechnik, Hannover, Germany, personal communication).
Unicompartmental knee replacement implants are widely used in humans (Koskinen and others 2007) but have not yet been developed for use in dogs. However, a femoral trochlear replacement (Patellar Groove Replacement; Kyon AG, Zurich, Switzerland) is currently under clinical investigation as a surgical option for managing dogs with advanced patella-femoral joint disease.
Total disc replacement
A commercial canine intervertebral disc (IVD) replacement has been developed and laboratory studies have confirmed that the implant is effective in restoring cervical IVD motion following discectomy (Adamo and others 2007). As in humans, the -proposed advantages of disc replacement over spinal fusion include preservation of spinal motion and a reduced risk of adjacent level disease (so-called domino lesions). However, disc replacement is known to be associated with a risk of potential complications such as implant migration, endplate subsidence, infection and implant wear (Salari and McAfee 2012). A clinical report on two dogs was recently published (Adamo 2011). Cervical IVD replacement appears to be technically feasible but much remains to be learned regarding both long-term results and the appropriate selection criteria for determining whether a dog should undergo IVD replacement rather than a discectomy/fusion procedure.
Total hock replacement
Given the incidence of hock osteochondrosis in dogs, it is not surprising that advanced osteoarthritis is a significant problem in this joint. While not ideal, the clinical results following talocrural arthrodesis are generally acceptable to most owners, with 12 of 13 dogs having good or excellent outcomes in a recent cases series (McKee and others 2004). Nevertheless, in select cases it may be appropriate to consider a motion-preserving procedure such as joint replacement, rather than fusion. The primary advantages of hock arthroplasty would be to preserve motion while also relieving pain associated with joint degeneration. As such, the indications for surgery would be similar to those for ankle joint replacement. To this author's knowledge, there has only been one case of hock arthroplasty to date (Dr. Noel Fitzpatrick, personal communication) and although the surgery was shown to be technically feasible, it is too early to draw any conclusions regarding the clinical outcome.
Total shoulder replacement
As compared with humans (Safran and Ianotti 2006), shoulder arthrodesis results in much less functional impairment in the dog, with only a relatively mild residual gait abnormality in dogs, most likely because the scapula accommodates by increasing its motion relative to the body wall (Dyce 1996). In a recent clinical report, shoulder arthrodesis resulted in 87·5% good to excellent functional results in dogs with shoulder instability (Pucheu and Duhautois 2008). Successful use of both shoulder hemiarthroplasty (Sparrow and Fitzpatrick 2012) and a total shoulder prosthesis have been reported (Fitzpatrick 2011), but there is currently no indication that a commercial implant system will become available in the near future.
Potential complications of TJR
Luxation is the most common short-term complication in both THR and TER. In THR, reported rates for luxation range from 1·1 to 8·5% of cases (Dyce and others 2000). The risk of luxation is determined by both implant design and surgical technique. A retrospective clinical study with the cemented BioMedtrix system recommended that the acetabular cup should be implanted at an angle of lateral opening of 35° and 45° in order to decrease the risk of dorsal luxation (Dyce and others 2000). A subsequent study on ventral luxation, which is seen less often than dorsal luxation after THR, demonstrated that short neck lengths are associated with an increased risk of luxation, while cup orientation was not a significant factor (Nelson and others 2007).
Aseptic loosening is the most common long-term complication following TJR surgery in humans. A retrospective review of 97 cemented THR identified aseptic loosening as a reason for revision surgery in 6% of cases (Bergh and others 2006) but a retrieval study performed at Colorado State University identified aseptic loosening (defined by a combination of radiography and mechanical testing) in 63% of implants (Skurla and others 2005) although lameness was not apparent in all cases. In cemented stems, failure at the stem-cement interface appears to precede failure at the cement-bone interface. It is likely that debris generated from abrasive wear between the metal stem and the cement mantle contribute to the development of an inflammatory response in the synovium (Day and others 1998) and the formation of the classic “interfacial membrane” at the cement-bone interface (Turner and others 1993).
Infection is a devastating complication of TJR. Risk factors for infection include prolonged operative time and a history of previous surgery on the joint (Conzemius and Vandervoort 2005). The use of antibiotic-loaded bone cement is recommended in all cases of cemented TJR (Lee and Kapatkin 2002). If sepsis does develop, treatment involves explantation of the prosthesis, debridement of the periprosthetic tissues (including removal of the cement mantle, if present) and a protracted course of systemic antimicrobial therapy. A technique for explanting the femoral component through a window in the lateral cortex has been described (Torres and Budsberg 2009) and may be particularly useful in failed cemented femoral stems. Revision of an infected TKR would be much more challenging, however, since removal of the implants would have to be accompanied by either arthrodesis or by replacement with a revision implant and antibiotic-loaded bone cement. A similar problem exists for the elbow, and to date there have not been any reports of successful management of infected TER or TKR in dogs.
Periprosthetic fractures are seen sporadically but can be extremely troublesome. Potential causes of fracture include trauma, intraoperative fissures (Liska 2004b) or the development of stress risers immediately distal to the tip of a femoral stem or at a notch in the cranial femoral cortex created during TKR (Liska and Doyle 2009). Many of these fractures can be managed very effectively using standard internal fixation techniques (Liska 2004b, Fitzpatrick and others 2011).
Ligamentous injuries are a particular concern in TKR and TER since the stability of both of these joints is critically dependent on the integrity of periarticular soft tissues such as the joint capsule and collateral ligaments. Collateral damage may be a result of intra-operative surgical error (especially during the resection of the proximal tibia or distal femur during TKR), or it may develop postoperatively as a consequence of abnormal joint kinematics, incorrect soft tissue tensioning or excessive postoperative activity (Liska and Doyle 2009). In our experience, intraoperative ligament tears are amenable to primary repair using heavy gauge suture, or the ligament can be replaced with a combination of fibre tape or fibre wire and suture anchors (Fig 7). In either case, use of a custom orthotic for the first six to eight weeks -postoperatively is recommended as a means of protecting the ligament repair.
Current controversies in TJR
The pathophysiology of aseptic loosening involves the development of an inflammatory response to particulate wear debris liberated from the surfaces of the metallic and polymeric implants (Harris 2001). In THR, the majority of the wear debris is generated through abrasive wear mechanisms; in contrast, wear in TKR more often involves adhesive wear mechanisms that result in the generation of a smaller number of much larger wear particles (Schmalzried and others 1997).
There has only been one published report on the systematic assessment of wear in canine THR (Skurla and James 2001). In that study, the authors found that volumetric UHMWPE wear rates in well-functioning acetabular components paired with either the modular BioMedtrix femoral component (6·9 mm3/year) or the monolithic Richards femoral component (8·3 mm3/year) were much lower than those seen in human post-mortem retrievals (with 35·0 mm3/year). Not surprisingly, the wear rate was significantly higher in cups that were obtained at revision surgery. While retrieval studies of this type cannot establish a causal link between wear and subsequent implant failure, there is a huge body of experimental evidence indicating that the presence of large amounts of wear debris at the cement-bone or bone-implant interface results in osteolysis and mechanical loosening (Shanbhag and others 1997). Since traditional wear simulators are not well suited to testing canine joint replacement implants, the best data on wear performance are likely to come from the analysis of well-functioning and failed implants (see later).
In a recent study on a series of canine TKR retrievals, we found evidence of subsurface failure, or delamination, of the UHMWPE in three of six implants (Rudinsky and others 2010). In human TKR, delamination has been associated with the development of excessive contact stresses within UHMWPE implants (Bartel and others 1986) and a similar mechanism is likely to be ongoing in canine TKR. In laboratory studies, the risk of delamination has been shown to be higher in implants with thinner UHMWPE inserts or with reduced contact area between the femoral and tibial components. In dogs, there are physical (size) limitations to the maximum thickness of the UHMWPE that can be implanted in the tibia (Bartel and others 1986). As a result, the most reasonable approach to limiting wear lies in accurate surgical implantation of the components, maintenance of optimal soft tissue balance, and use of semi-constrained implant designs that distribute femorotibial loading over an increased area of surface contact.
Rational selection of cemented versus cementless fixation
Established advantages of cementless fixation include reduced operative time, elimination of PMMA as a source of particulate debris, elimination of monomeric MMA as an airborne hazard to operating theatre personnel, and a decreased risk of thermal injury to bone.
Cementless THR appears to be extremely successful, as long as appropriate cases are selected for surgery. Use of undersized stems (Rashmir-Raven and others 1992), or implantation of cementless implants in femora with poor bone quality, results in an increased risk of implant subsidence and failure. Dogs with suboptimal bone quality or a stove-pipe femoral morphology (Rashmir-Raven and others 1992) may be better suited to hybrid THR with a cemented femoral component and a cementless acetabular component (Gemmill and others 2011). It should be noted that definitive clinical evidence supporting the superiority of either cemented or cementless fixation is currently lacking.
In human TKR, the clinical results of cementless tibial fixation are not as good as those seen with cemented tibial fixation (Robertsson and others 2001). However, one must be cautious in extrapolating human clinical data to dogs. There is a large body of preclinical data indicating that cementless fixation is feasible in the dog, and the author has recently confirmed bone ingrowth and effective cementless fixation with two different forms of porous-coated, metal-backed tibial components (Allen and others 2007, 2010). Cementless fixation also appears to be feasible in the TATE total elbow. As is the case with THR, the relative merits of cemented versus cementless fixation for the elbow or knee can only be determined through prospective randomised clinical trials.
Revision and custom TJR solutions
Revision options for canine THR have been reported, including revision of failed CFX implants with a second cemented implant (Edwards and others 1997), conversion of a failed Helica hip to a cementless Kyon hip (Andreoni and others 2010) and revision of failed cemented CFX implants to cementless BFX implants (Torres and Budsberg 2009) (Fig 8). Outcomes after revision surgery for aseptic loosening or luxation are generally good to excellent.
Revision options for canine TKR cases have until recently been very limited. In principle, implants that fail because of aseptic loosening can be removed and replaced with long stemmed implants that bypass regions with bone loss and permit cemented or cementless fixation in healthy distant bone stock. Implants that become infected are much harder to manage; in humans, a one- or two-step revision procedure is employed and infection is eradicated in less than 90% of cases (Westrich and others 2010), although long-term clinical function is often suboptimal. However, the cost of a two-stage revision can be prohibitive in dogs. Palliative management with antibiotic-impregnated bone cement beads may be attempted, but most dogs with infected TKR are ultimately managed with arthrodesis, amputation or euthanasia.
In the last couple of years, there has been renewed interest in the development of revision solutions for canine TKR, as well as in the development of custom implants that can be used to manage animals with significant knee joint instability (e.g. following stifle joint derangement, or recurrent medial collateral rupture) or dogs with malignant lesions in the distal femur or proximal tibial. Although case series are not available, the isolated case reports indicate that the use of hinged implants is feasible in these complex cases (Fitzpatrick and Asher 2010). However, data on the long-term survival of these implants must be collected before any conclusions can be drawn regarding the potential utility of these revision implants.
Perspectives on future developments in TJR
Objective assessment of TJR
In this era of evidence-based medicine, it is somewhat alarming to find that there is almost no evidentiary proof for the value of TJR compared with other surgical or non-surgical options in the management of advanced DJD in dogs. In a recent “Where's the Evidence” article in the Journal of the American Veterinary Medical Association, Franklin and Cook found only one published report of a head-to-head comparison of THR versus femoral head and neck excision (Dueland and others 1977) and concluded that “until these types of studies are completed, owners cannot truly perform a valid cost-benefit analysis and make a fully informed decision regarding their pet's surgery” (Franklin and Cook 2011).
While prospective, randomised controlled clinical trials can be complicated and expensive to run, it is critically important that the profession and the veterinary implant manufacturers grasp this nettle and move forward. Developments in surgical technique and implant design should be based on an established, rather than theoretical, need for improvement. The first step in this process is the identification and quantification of the problem, and this demands a comprehensive assessment of clinical outcomes with the current generation of implant designs (Iwata and others 2008). In human medicine, centralised joint registries have been established to track indications for surgery, the choice of implant, the nature of any operative or postoperative complications, and the patient's functional performance after surgery (Knutson and Robertsson 2010). A similar approach has been initiated in the United Kingdom, with the introduction of the British Veterinary Orthopaedic Association canine total hip registry (http://caninehipreplacement.org), hosted by the University of Liverpool (Innes 2012). Ideally, the demographic data collected through such a registry would be combined with either subjective data on clinical outcomes, using a validated survey instrument (Hudson and others 2004), or with objective measures of limb loading collected through either force plate or pressure walkway analysis (Dueland and others 1977, Massat and Vasseur 1994, Budsberg and others 1996). Large, multi-centre studies with objective outcome measures provide the only rational solution to questions about the ethical and financial viability of advanced medical and surgical procedures, and surgeons offering these procedures should be encouraged to help to develop the evidence necessary to support their use in clinical patients.
The role of retrieval studies
Although one commonly considers retrieval studies in the context of troubleshooting complications with a given implant system, there is tremendous value to be gained through more widespread introduction of a retrieval program. The authors group has recently established an implant retrieval program at The Ohio State University, the goal of which is to provide a technical resource for surgeons and implant manufacturers who are interested in better understanding the biological and mechanical performance of implants retrieved either during a revision -surgery or at the time of euthanasia/death of the animal. Previous studies with THR have demonstrated that an absence of overt clinical signs of lameness does not preclude evidence of implant loosening (Skurla and others 2005), and while the relatively short lifespans of small animal patients mean that we may not see overt implant failure in every case, the lessons learned from retrieval analysis or large-scale clinical studies can certainly inform our decisions with regard to future refinements in implant design and surgical technique (Lemons and others 2010).
Over the last 40 years, TJR has matured into an established treatment option for managing small animals with advanced joint disease. Despite the current economic conditions, many clients are interested in pursuing total hip, knee or elbow replacement for their pets. As the market for these procedures grows and new implants are developed in response to this demand, it is important to retain a healthy skepticism about new implant designs until there are objective clinical data to support their use. Clinical results for many of the existing implants are good to excellent, so any modifications to existing implant designs should be made on the basis of addressing clinical problems or technical limitations that have been identified through retrieval analysis. Implant manufacturers are doing an excellent job of providing training and certification in the use of their implants, and this trend should be encouraged. At the same time, surgeons who perform TJR and who have to make choices between different implant designs should be pressing manufacturers to provide objective data to support the safety, efficacy and longevity of their implants.
The author acknowledges the invaluable mentorship of John Houlton, former University Surgeon at the University of Cambridge, who first inspired my interest in comparative orthopaedics.
Conflict of interest
The author is a consultant for the Canine Total Knee program and receives consulting fees from BioMedtrix, LLC.