 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 2004 The Journal of Bone and Joint Surgery, Inc. What's New in Adult Reconstructive Knee Surgery1 New Mexico Orthopaedics, New Mexico Center for Joint Replacement, 201 Cedar S.E., Suite 6600, Albuquerque, NM 87106. E-mail address for M.J. Archibeck: archibeckmj{at}ortholink.net
Specialty Update has been developed in collaboration with the Council of
Musculoskeletal Specialty Societies (COMSS) of the American Academy of
Orthopaedic Surgeons.
The purpose of this update is to discuss, in summary fashion, topics presented at selected orthopaedic meetings and published in related orthopaedic journals during 2003. The sources for this review are articles published in The Journal of Bone and Joint Surgery (American edition) and The Journal of Arthroplasty. The podium presentations mentioned in this article include those given at the annual meeting of the American Academy of Orthopaedic Surgeons (held in New Orleans, Louisiana, on February 5 through 9, 2003), on Specialty Day at the meeting of the Knee Society (held in New Orleans, Louisiana, on February 8, 2003), at the interim meeting of the Knee Society (held in Chicago, Illinois, on August 27 through 29, 2003), and at the annual meeting of the American Association of Hip and Knee Surgeons (held in Dallas, Texas, on October 31 through November 2, 2003).
Although total knee replacement has been very effective for alleviating pain and improving function in patients of all ages, other approaches remain the initial treatment for many patients with knee arthritis. Leopold et al., in a prospective, randomized study, compared intra-articular corticosteroids with hyaluronic acid injections and found no differences in pain or function at six months, with significantly less response in women relative to men in both groups (p < 0.05). Dervin et al. prospectively evaluated 126 patients who underwent arthroscopic knee débridement for the treatment of symptomatic osteoarthritis and found that 44% of the patients had substantial reduction in pain according to the WOMAC pain scale at two years of follow-up. Shasha et al., in a study of sixty-five knees, reviewed the results associated with the use of fresh osteochondral allografts for the treatment of posttraumatic tibial plateau defects and found good clinical results and a survival rate of 80% at ten years. Several investigators discussed the role of osteotomy in the treatment of the arthritic knee. Sprenger and Doerzbacher1 reviewed the results of seventy-six valgus-producing proximal tibial osteotomies after a mean duration of followup of 10.8 years and found a ten-year survival rate of 74% with conversion to arthroplasty as the end point. This ten-year survival rate improved to 90% when the radiographic valgus angle at one year was 8° to 16°. Koshino et al. evaluated the results of twenty-one medial opening-wedge proximal tibial valgus osteotomies after a mean duration of follow-up of 6.6 years and found pain relief and improved walking ability in all patients, with the mean Knee Society score improving from 60 to 94 points at the time of the last follow-up. Chiang et al. evaluated the clinical results as well as the results of gait analysis for twenty patients who had been managed with a medial opening-wedge proximal tibial osteotomy. Clinical improvement was noted in all patients and gait analysis demonstrated improvement in terms of both the postoperative vertical ground-reaction force and the adduction moment.
With the development of less-invasive techniques and improvement in the reported rates of survival, unicompartmental knee replacement has become increasingly popular. Jenny and Boeri found that the use of intramedullary femoral instrumentation and extramedullary tibial instrumentation was associated with improved accuracy in terms of unicompartmental implant placement when compared with older extramedullary techniques. The rate of implant survival, however, was unaltered after five years. Hernigou et al. reviewed the results of 156 medial unicompartmental knee replacements that had been performed with cement and found that postoperative valgus alignment was a risk factor for adjacent-compartment degeneration and revision whereas severe undercorrection was a risk factor for tibial component loosening. The best results were seen in patients in whom the hipknee-ankle angle on radiographs of the whole limb was 171° to 179°. Those data support the technical dictum that slight undercorrection of the preoperative deformity is the desired postoperative alignment. Several investigators reported on the clinical results of unicompartmental knee replacement. Keblish and Briard evaluated the results of 153 mobile-bearing unicompartmental knee replacements (72% of which had been performed without cement) after a mean duration of follow-up of ten years and found an 84% rate of good and excellent results and a 15.7% rate of reoperation (twenty-four unicompartmental knee replacements). Seven reoperations (including three for tibial component loosening, one for tibial plateau fracture, and three for component malposition) were performed within the first two years, and seventeen late failures occurred as a result of wear and/or bearing fracture. The rate of implant survival was 84% at ten years. Gioe evaluated the results of 516 unicompartmental knee replacements that had been followed in a community-based implant registry and reported a revision rate of 7.6% (thirty-nine of 516) and a ten-year survival rate of 88.6%. The author reported that 51% of the revisions had been performed because of adjacent-compartment degeneration, 26% had been performed because of component loosening, and 21% had been performed because of polyethylene wear. Argenson et al. reviewed the results of 160 unicompartmental knee replacements that had been performed with cement and reported five revisions (including three that had been performed for adjacent-compartment degeneration and two that had been performed for polyethylene wear) after a mean duration of follow-up of 5.5 years. The ten-year implant survival rate was 94% ± 3%. Berger et al. evaluated the results of sixty-two consecutive unicompartmental knee replacements after a minimum duration of follow-up of ten years and reported that two revisions had been performed because of adjacent-compartment degeneration. The implant-survival rate was 98% at ten years and 96% at fourteen years. Pennington et al. evaluated the results of forty-six unicompartmental knee replacements in patients who were sixty years of age or younger and reported three revisions (two of which were performed because of polyethylene wear and one of which was performed because of pain and the presence of radiolucent lines) after a mean duration of follow-up of eleven years. The implant-survival rate was 92% at eleven years. Tria evaluated the results of 273 minimally invasive unicompartmental knee replacements that had been performed during the previous four years and reported four failures (including one that was due to femoral component loosening, one that was due to patellar dislocation, and two that were due to arthritic progression). Polyethylene wear has been found to be a common mode of failure after unicompartmental knee replacement. Collier et al. reported on 100 unicompartmental knee replacements performed with polyethylene components that had been gamma sterilized in air and had a median shelf life of 1.7 years. The rate of failure was 25% (twelve of forty-eight) among implants that had a shelf life that was above the median and only 4% (two of fifty) among those that had a shelf life that was below the median. A unique complication of unicompartmental knee replacement is that of tibial plateau stress fracture. Brumby reported on four such fractures that occurred in association with the use of four guide-pins that had been placed to secure the tibial cutting guide. Yang et al. reported on two medial tibial stress fractures that occurred following minimally invasive unicompartmental knee replacement.
While the clinical results of total knee arthroplasty continue to be excellent, debates regarding the superiority of various fixation techniques and differing designs continue. Forster performed a survival analysis that was based on the compiled data from sixteen articles (involving a total of 5950 knees) and found that posterior stabilization or the use of a metal-backed tibial component does not improve the longevity of primary cemented fixed-bearing condylar-type total knee replacements. Rand et al. performed a survivorship analysis of 11,606 total knee replacements and found survival rates of 91%, 84%, and 78% at ten, fifteen, and twenty years, respectively. The implant-survival rate was better for patients who were more than seventy years old as compared with those who were less than fifty years old (94% compared with 83%), patients who had inflammatory arthritis (95% compared with 90% in patients with osteoarthritis), patients who had posterior cruciate ligament retention (91% compared with 76% in posterior stabilized knees), patients who had nonmodular tibial components (92% with metal-backed nonmodular components, 97% with all polyethylene components, and 90% with modular components), and patients who had cement fixation (92% compared with 61% in patients with cementless fixation) (p < 0.005 for all comparisons)2. Gill and Joshi reported on 1033 cemented posterior cruciate ligament-retaining total knee replacements and found the probability of survival to be 95% at fifteen years, 89% at twenty years, and 83% at twenty-four years, with inferior implant-survival rates in patients with osteoarthritis (82%) and in female patients (81%) at twenty-four years. Meding et al. evaluated the results of 220 posterior cruciate ligament-retaining total knee arthroplasties that had been performed in 148 patients with rheumatoid arthritis and found a 2.4% rate of deep infection, no femoral revisions, and a 1.4% rate of tibial revision (with one revision performed because of loosening and two performed because of instability) after a mean duration of follow-up of ten years. Fixation techniques remain controversial as well. Goldberg et al. evaluated the results of 124 consecutive cementless total knee arthroplasties after a minimum duration of followup of fourteen years and noted only one case of tibial revision but a 21% rate of tibial osteolysis. Mahoney et al. evaluated the results of 183 total knee arthroplasties that had been performed with use of posterior stabilized modular implants that had been inserted with cement (49%) or with hybrid fixation (51%) and found that nine femoral components (including two cemented implants and seven hybrid implants) had been revised because of loosening and that eight polyethylene inserts had been exchanged after eleven to thirteen years of follow-up. Osteolysis was more pronounced in knees with cementless femoral components, but it did not seem to affect survival at this interval. Illgen et al. reviewed the results associated with 112 hybrid total knee replacements (involving sixty-seven PCA implants and forty-five Duracon implants [Howmedica, Rutherford, New Jersey]) and reported a 4.5% revision rate (with four revisions performed because of complications associated with a metal-backed patellar component and one revision performed because of infection) after a mean duration of follow-up of ten years. There were no cases of aseptic loosening. The role of supplemental hydroxyapatite coating on femoral components was addressed by Murty and Scott, who reviewed thirty-six such implants after a minimum duration of follow-up of ten years and identified two instances of femoral loosening (in one patient). As a result of recent reports on tibial component back-side wear, the role of nonmodular tibial components has gained some attention. Faris et al. evaluated 536 AGC total knee replacements (Biomet, Warsaw, Indiana) that had an allpolyethylene tibial component and noted a relatively poor survival rate (68%) at ten years. Fetzer et al. evaluated the results of 101 press-fit condylar total knee arthroplasties (PFC; Johnson and Johnson, Raynham, Massachusetts) that had been performed with use of a modular metal-backed tibial component and noted only one revision and one instance of radiographic failure after a mean duration of follow-up of 10.5 years, with an implant-survival rate of 93% at twelve years. One potential method with which to address the issue of backside wear is the use of a mobile-bearing total knee replacement. Several reports on such designs were presented or published in 2003. Bhan and Malhotra evaluated the results associated with thirty-one rotating platform total knee replacements (LCS; DePuy, Warsaw, Indiana) in knees with "deformity" and reported high complication rates (including a 2% rate of dislocation, a 4% rate of anteroposterior instability, and a 10% rate of subluxation) and an overall reoperation rate of 10% after a mean duration of follow-up of 4.5 years. They concluded that the LCS is unsuitable for severely deformed knees. Nakamura et al. evaluated the results of eighty-two consecutive mobile-bearing total knee arthroplasties that had been performed without cement and compared them with the results of a subsequent series of seventy-six mobile-bearing total knee arthroplasties that had been performed with cement. After a minimum duration of follow-up of two years, six (8%) of seventy-three cementless tibial implants and zero of sixty-six cemented tibial implants had been revised because of the failure of ingrowth (p < 0.05), thereby refuting the hypothesis that mobile-bearing total knee arthroplasties impart the advantage of reliable tibial bone ingrowth. Morgan-Jones et al. evaluated seventy-five meniscal-bearing total knee arthroplasties (Motus; Osteo, Selzach, Switzerland) and noted no instances of revision and no bearing complications after a mean duration of follow-up of 2.5 years.
While minimally invasive total knee replacement and computer-assisted orthopaedic surgery are still relatively new, several reports were presented or published on these topics. Tria evaluated the results of the first seventy minimally invasive total knee arthroplasties that he had performed and reported that this technique was associated with decreased blood loss, shorter hospital stays, increased range of motion, and similar implantation accuracy when compared with standard total knee replacement. Laskin evaluated the results of fifty-one total knee arthroplasties that had been performed through a mini-mid vastus approach and found only two cases in which the incision needed to be extended. This approach was associated with a mean operative time of sixty-three minutes, a shorter time to straight leg-raising compared with that following procedures performed through a standard incision (1.2 days compared with three days), an improved range of motion prior to discharge, and less use of pain medication when compared with that after fifty-one previous primary total knee arthroplasties. Laskin also noted excellent radiographic positioning of the implants (within ±2° of the intended position). Scuderi proposed some important potential contraindications to minimally invasive total knee arthroplasty, including substantial deformity (e.g., varus-valgus deformity and flexion contracture), poor flexion, a wide femur, and patella baja. Stulberg et al., with use of a navigation system, found that errors are most likely to occur when pins are placed manually, when blocks are placed on osteoporotic bone, when sagittal cuts of the femur or tibia are made, and when the tibial component is impacted onto osteoporotic bone. Wasielewski and Komistek reported on thirty-eight patients who underwent posterior cruciate ligament-sacrificing total knee replacement (LCS; DePuy). At the time of trial reduction, an instrumented tibial insert was used to measure compartment pressures and ligament balance. The authors found this system to be more sensitive than "surgeon feel," suggesting that it may be helpful for achieving ligamentous balance. Perlick et al., in a randomized study of eighty patients, compared computer-navigated computed tomography-based implantation with conventional implantation and noted significantly better mechanical femorotibial alignment in the study group; specifically, the implant was positioned within ±2° of the intended position in 73% of the patients in the study group compared with 53% of those in the control group. Jenny et al. compared 555 total knee arthroplasties that had been performed with use of a non-image-based navigation system with 266 historic controls improved mechanical alignment in the study group; specifically, the implant was positioned within ±2° of the intended position in 79% of the patients in the study group compared with 61% of those in the control group) (p < 0.01).
Several reports were presented on the topic of patient management in the perioperative period. Rosen et al. compared the results of eighty-five total knee arthroplasties during which tourniquet release had been performed prior to closure with those of sixty-three total knee arthroplasties during which release had been performed following dressing application and found no differences in terms of early drain output, decrease in hematocrit, transfusion rates, or wound complication rates (p > 0.05). DeWal et al. reviewed the factors influencing the duration of postoperative wound drainage in a study of 2001 primary total knee arthroplasties and found that body-mass index correlated positively with drain output and that the use of low-molecular heparin was associated with a prolonged time to a dry wound. There were several reports on issues related to surgical technique as well. Tanaka et al. evaluated the effect of infrapatellar fat pad resection in a study of 120 patients with rheumatoid arthritis and found shortening of the patellar tendon, reduced motion, and anterior ache following infrapatellar synovectomy. Matsuda et al. compared three different techniques for tibial preparation in a study of thirty total knee arthroplasties and found that the use of an original template connecting the center of the cut tibial plateau to the center of the talar dome was more accurate than the use of the tibial shaft as the alignment guide. Healy et al. evaluated mesh expansion release of the lateral retinaculum in a study of thirteen total knee arthroplasties and reported good results and no complications at two years. Ranawat reported on seventy-one total knee arthroplasties in patients with a preoperative valgus deformity of >10° who were managed with an inside-out release of the posterolateral capsule and pie-crusting of the iliotibial band, with no late instability or complications related to the release.
The most appropriate form of prophylaxis against thromboembolic disease following total knee arthroplasty remains controversial. Lachiewicz reported the results of a prospective, randomized study in which 206 patients who were managed with an asymmetric calf-compression device were compared with 217 patients who were managed with a circumferential calf-compressive device. Both groups also received aspirin. The authors found a significantly lower prevalence of deep venous thrombosis in the group that was managed with the asymmetric compression device (6.9% compared with 15%; p = 0.005). One fatal pulmonary embolus was identified in the group that was managed with the circumferential calf-compression device. Westrich et al. studied 229 patients who had been randomized to treatment with aspirin and calf pumps or with low-molecular-weight heparin and calf pumps and reported a similar prevalence of deep venous thrombosis on the third through the fifth postoperative day (18% compared with 16%) and a lower prevalence of deep venous thrombosis in the heparin group at four weeks postoperatively (5.5% compared with 1.3%). Reitman et al. reported the results of a study of 1308 total knee arthroplasties (964 procedures) in which the patients were managed with intraoperative heparin (1500 units), hypotensive epidural anesthesia, pneumatic compression boots, and aspirin for prophylaxis against deep venous thrombosis. Routine venous Doppler screening was performed at the time of discharge. The authors reported a 4% rate of deep venous thrombosis (including a 1% rate of proximal deep venous thrombosis and a 3% rate of distal deep venous thrombosis). Francis et al. reported the results of a randomized, double-blind trial of 2301 patients in which ximelagatran (administered at a dose of 36 mg twice daily) was compared with warfarin for the prevention of deep venous thrombosis. Ximelagatran demonstrated superior efficacy with regard to the prevention of deep venous thrombosis (prevalence, 20.3% compared with 27.6%; p = 0.003). There was no difference between the groups with regard to major bleeding complications, wound characteristics, pulmonary embolus, or death. The use of anticoagulants in the perioperative period is not without risk. Sachs et al.3 compared a cohort of 785 patients who were treated without any form of thromboprophylaxis with a cohort of 957 patients who were treated with six weeks of low-dose warfarin. The control group had an overall complication rate of 2.2%, no deaths due to pulmonary embolism, and a total death rate of 0.2%. In comparison, the warfarin group had an overall complication rate of 4.7%, no deaths due to pulmonary embolism, a total death rate of 0.1%, and twice the infection rate of the control group. The authors concluded that anticoagulation may do more harm than good. Keays et al. compared the effects of two anticoagulants (aspirin and enoxaparin) on the postoperative range of motion and found that the patients in the aspirin group reached range-of-motion milestones (90°, 100°, and 110°) earlier than did those in the enoxaparin group (p < 0.001). The authors noted, however, that there was no significant difference between the groups with regard to the range of motion at fifteen months. Berend et al., in a study of patients treated with aspirin and mechanical devices, found that the odds ratio for deep venous thrombosis increased to 5.5 when patients who had a postoperative ileus following total knee arthroplasty were compared with patients who did not have an ileus (p = 0.0036).
Several authors reviewed the results of primary total knee arthroplasty in young and elderly patients. Gill and Joshi, in a study of 106 total knee arthroplasties in patients who were fifty-five years old or less (mean age, forty-eight years) and who had a minimum duration of follow-up of ten years, reported a failure rate of 6.6% and a fifteen-year survival rate of 90%. Jordan et al., in a report on sixty-seven cementless total knee arthroplasties (LCS; DePuy) that had been performed in patients who were sixty years old or younger (mean age, fifty-six years), reported a 22% revision rate. Thirteen of the fifteen revisions were performed because of wear or bearing fracture. The thirteen-year survival rate was only 69% with revision for any reason as the end point. Addressing the other age extreme, Berend et al. evaluated the mean 2.5-year results of 101 arthroplasties (including forty-five total knee arthroplasties and fifty-six total hip arthroplasties) in patients who were eighty-nine years of age or older. The authors reported a 31% mortality rate during the follow-up period and a 14% rate of medical complications. The survival rate was reported to be at least equal to that in the age-matched population, and the authors concluded that total joint replacement can be an excellent option for these patients. Joshi et al. evaluated the results of 110 total knee arthroplasties in ninety patients who were eighty years of age or older. All living patients had a minimum duration of follow-up of ten years. The authors reported that 96% of all patients had complete pain relief and that none of the implants failed because of loosening. Several authors reported on the results of total knee arthroplasty in patients with specific preoperative diagnoses. Parvizi evaluated the results of forty total knee arthroplasties in twenty-nine patients with Charcot arthropathy after a mean duration of follow-up of eight years. Ligamentous instability necessitated the use of long-stem components in twenty-seven knees and a rotating-hinge prosthesis in five. There were six reoperations (two for fracture, two for component loosening, one for instability, and one for infection). Ito et al. performed a fifteen-year follow-up study of thirty-six total knee arthroplasties in patients with rheumatoid arthritis and found a 77.7% rate of good or excellent results and a 93.7% rate of survival at fifteen years with revision as the end point. Parvizi et al.4, in a report on twenty-five total knee arthroplasties in thirteen patients with juvenile rheumatoid arthritis (mean age, seventeen years) who had been followed for a mean of 10.7 years, noted marked improvement in the Knee Society pain score (from 28 to 88 points) and modest improvement in the function score (from 15 to 39 points). The authors reported one revision for loosening, two polyethylene exchanges, and four additional reoperations. Nau et al. reported good clinical results in a series of six elderly women with severe osteopenia and periarticular fractures who had been managed with a primary total knee arthroplasty with use of a hinged prosthesis (five patients) or an unconstrained prosthesis (one patient) followed by immediate mobilization and full weight-bearing. Parvizi et al. evaluated the results of twenty-one arthroplasties in fifteen patients with the human immunodeficiency virus after a mean duration of follow-up of 10.2 years and identified six deaths, six deep infections, and thirteen reoperations—an alarmingly high rate of complications. Helenius et al. evaluated the results of twenty-one total knee arthroplasties in fourteen patients with diastrophic dysplasia after a mean duration of follow-up of 3.4 years and found that two patients had required a distal femoral corrective osteotomy for angulation; there were six complications and no revisions. The John Insall Award was given to Schrader and Morrey for their report on eighty-three total knee arthroplasties in sixty-three patients with lymphedema who had been followed for a minimum of two years. The authors reported good clinical results and a total complication rate of 31%; the complications included ten superficial infections, six deep infections, and three deep venous thromboses. Silva and Luck reviewed the results of ninety-one total knee arthroplasties in sixty-eight patients with hemophilia (68% of whom were positive for the human immunodeficiency virus) and reported eight revisions (including five for infection, two for component loosening, and one for recurrent hemarthrosis) after a mean duration of follow-up of 7.5 years. Meding et al. reviewed the results of 363 primary total knee arthroplasties in 291 patients with diabetes mellitus and found a 1.3% rate of deep infection and a 3.7% rate of revision after a mean duration of follow-up of fifty-two months. Nelson et al. reviewed the results of eleven total knee arthroplasties that had been performed after a varus osteotomy of the distal part of the femur. A constrained prosthesis was used in five of the eleven knees. The authors reported that the procedure was technically demanding and that it demonstrated inferior results when compared with those of primary total knee arthroplasty in patients without a prior osteotomy. Weiss et al. reviewed the results of sixty-two total knee arthroplasties in patients with a previous tibial plateau fracture. The authors reported good clinical results after a mean duration of follow-up of 4.7 years, with six intraoperative complications (prevalence, 10%), sixteen postoperative complications (prevalence, 26%), and thirteen reoperations (prevalence, 21%). The reoperations included five manipulations, three wound revisions, and five component revisions.
The controversy regarding patellar resurfacing continues. Mayman et al. performed a prospective study of 100 total knee arthroplasties in patients who were randomized to patellar resurfacing or nonresurfacing. All patients were managed with a single prosthesis that featured an anatomically designed patellofemoral articulation (AMK; DePuy) The authors reported an 80% rate of extreme satisfaction in the resurfacing group and a 48% rate of extreme satisfaction in the nonresurfacing group after eight to ten years of follow-up. Waters and Bentley5 performed a prospective, randomized study of 514 total knee arthroplasties performed with the Press-Fit Condylar prosthesis (Johnson and Johnson). After a mean duration of follow-up of 5.3 years, the prevalence of anterior knee pain was 25% in the nonresurfacing group and 5% in the resurfacing group (p < 0.0001). Eleven of the unresurfaced patellae were subsequently resurfaced; ten of these eleven secondary resurfacing procedures were successful. There also has been renewed interest in isolated patellofemoral resurfacing for the treatment of symptoms limited to the patellofemoral joint. Ackroyd et al., in a study of ninety-five Avon patellofemoral arthroplasties, reported that the median Bartlett patella score had improved from 10 (of 30) to 26 (of 30) after two to five years of follow-up. The authors also reported a 1% rate of malalignment, a 7% rate of disease progression, and a 6% rate of conversion to total knee arthroplasty.
The Mark Coventry Award was given to Dennis et al. for their multicenter analysis of in vivo kinematics following 811 total knee arthroplasties. The authors found that patients who had been managed with mobile-bearing posterior-stabilized or fixed-bearing posterior-stabilized total knee arthroplasty experienced similar kinematic patterns as did those who had been managed with mobile-bearing posterior cruciateretaining and fixed-bearing posterior cruciate-retaining total knee arthroplasty. Posterior cruciate ligament-retaining designs are associated with a greater prevalence of paradoxical anterior femoral translation of both condyles. Komistek et al. used in vivo fluoroscopy to evaluate polyethylene bearing motion in ten subjects who had a posterior-stabilized mobile-bearing total knee replacement and found an average of 2.6° of polyethylene rotation relative to the tibia from full extension to flexion during weight-bearing. The kinematics of posterior cruciate ligament-retaining total knee arthroplasty were evaluated in two studies. In the study by Conditt et al., eight cadaveric specimens were mounted in a simulator and were flexed from 0° to 100° with varying amounts of posterior slope. The authors found that femoral rollback changed dramatically with posterior slope and that reduction of the posterior slope recreated rollback throughout the entire flexion arc. In the study by Bergin et al., in vivo kinematics were evaluated as twenty subjects with a posterior cruciate ligament-retaining total knee replacement performed deep-knee-bend activities. The authors found that nineteen subjects had posterior femoral rollback of the lateral femoral condyle (mean, 3.9 mm) and thirteen had posterior femoral rollback of the medial femoral condyle (mean, 3.1 mm), findings that contrasted with those of previous in vivo studies of posterior cruciate ligament-retaining designs.
Many reports dealt with the treatment of complications following total knee replacement. Debate continues with regard to the indications for simultaneous bilateral total knee replacement. Bullock et al. compared the results of 514 unilateral total knee arthroplasties with those of 255 simultaneous bilateral arthroplasties and found that the prevalence of myocardial infarction, the rate of postoperative confusion, and the need for intensive monitoring were higher in the bilateral group. However, the rates of infection, deep venous thrombosis, pulmonary embolus, and mortality were similar in both groups. Ritter et al. reviewed the results of 2050 simultaneous bilateral, 1796 unilateral, and 152 staged bilateral total knee arthroplasties and found a significantly higher rate of deep venous thrombosis in the simultaneous group (0.9% in the bilateral group compared with 0.3% in the unilateral group; p = 0.0326) but no differences with regard to the rates of prosthetic failure, cardiac complications, or death. Leonard et al. compared the results of ninety-two simultaneous total knee arthroplasties with those of ninety-two matched unilateral total knee arthroplasties. There was one death in the bilateral group. Cardiopulmonary complications occurred in six patients in the bilateral group, compared with two patients in the unilateral group. Bezwada et al. compared a group of seventy-five patients (150 knees) who underwent simultaneous revision and contralateral primary total knee replacement with a control group who underwent revision only and found the simultaneous procedures to be safe. Nunley and Lachiewicz reported that the ninety-day mortality rate in their study of total knee arthroplasties that had been performed by a single surgeon in a medium-volume practice was 0.33% (two of 610), with no difference between unilateral and simultaneous bilateral procedures with regard to the mortality rate at one or two years (p = 0.55). Gill et al. reported that the ninety-day mortality rate in their study of total knee arthroplasties that had been performed in a private-practice, nonteaching setting was 0.46% (fourteen of 3048), with increasing age and cardiovascular comorbidities being risk factors. Stiffness following total knee replacement remains a difficult problem to treat. Ritter et al. studied the results of 4727 total knee arthroplasties (in 3066 patients) and found preoperative range of motion to be the principal predictive factor. Lotke et al. found the rate of stiffness (defined as a flexion contracture of >15° and/or an arc of motion of <60°) to be 1.3%. Those authors reviewed the results of fifty-six revisions that had been performed for the treatment of stiffness and found a modest mean improvement in the arc of motion from 55° to 82° postoperatively. Lo et al. reported on two cases of knee stiffness secondary to the use of an oversized femoral component and noted substantial improvement after revision. Bush-Joseph et al., in a study of fifteen consecutive patients with isolated arthrofibrosis following total knee arthroplasty who had been managed with arthroscopic débridement and manipulation, reported an improvement in extension from 13° to 4° and an improvement in flexion from 60° to 80°. There was one infection, and 27% of the patients needed a subsequent revision. Toyoda et al., in a report on sixty-three total knee replacements, found that twenty-five knees (40%) had postoperative heterotopic ossification that did not affect range of motion. None of the knees in the control group had heterotopic ossification, suggesting that heterotopic ossification after total knee arthroplasty is a self-limiting condition. Brander et al., in their Chitranjan Ranawat Award-winning paper that was presented at the open meeting of the Knee Society, found that postoperative pain was significantly associated with preoperative depression (p = 0.001) and anxiety (p < 0.001) and was not associated with gender, age, weight, or type of anesthesia. As expected, patients with more postoperative pain required more physical therapy and had a higher rate of manipulation. Extensor mechanism disruption following total knee arthroplasty was discussed in several reports. Dobbs et al., in a report on thirty patients who had a quadriceps tendon rupture, noted a high rate of complications (eleven of twenty-four) among patients who had been managed surgically; these complications included four reruptures and three infections. Reis et al. compared six patients who had been managed with an Achilles-tendon allograft with ten patients who had been managed with a medial gastrocnemius rotational flap for the treatment of chronic extensor mechanism failure and found a 50% rate of failure in the allograft group and no failures in the rotational flap group. Sierra et al. reported that the prevalence of above-the-knee amputation following total knee arthroplasty at the Mayo Clinic was 0.36% (sixty-seven of 18,443) and noted that most (63%) of the amputations were performed for reasons that were not attributable to the total knee arthroplasty. They found that the functional outcome of amputation was poor, with infrequent use of a prosthesis and with patients seldom achieving functional independence.
Infection is one of the most dreaded complications following total knee replacement. Joshi and Gill reviewed 3048 total knee arthroplasties and found that the rate of early superficial infection was 0.8%, the rate of deep infection within ninety days was 0.13%, and the rate of deep late infection was 0.38%. Tang et al. compared the use of three doses of cefuroxime (215 arthroplasties) with a single preoperative dose of cefazolin (1152 arthroplasties) and found similar efficacies in the prevention of deep infection. Mason et al.6, in a study of 440 knees, reported on the value of white blood-cell counts before revision total knee arthroplasty and found that a white blood-cell count of >2500 cells/mm3 (>2.5 x 109/L) with >60% polymorphonuclear cells was highly suggestive of infection. McPherson et al., in a study of fifty-one patients who had a chronic infection at the site of a total knee arthroplasty, found that, at the time of explantation and reimplantation, the culture samples that were most likely to yield a positive result were those obtained from the tibial and/or femoral medullary canals, supporting the concept that the medullary canal should be débrided and lavaged at the time of explantation and reimplantation. The treatment of infection following total knee replacement was the topic of several reports. Nazarian et al., in a study of fourteen patients in whom septic knee arthritis or osteomyelitis was treated with resection arthroplasty, placement of an antibiotic spacer, and subsequent total knee arthroplasty, reported no recurrent infections at the time of the last followup (at a mean of 4.5 years). Deirmengian et al., in a study of thirty-one patients who had an acute infection with gram-positive bacteria at the site of a total knee arthroplasty, reported that open débridement, component retention, and intravenous administration of antibiotics was associated with a 35% cure rate. When Staphylococcus aureus was the pathogen, the rate of successful treatment was only 8%, suggesting that immediate component removal may be the preferred treatment in cases of S. aureus infection. Berry et al., in a report on ninety-four knees that underwent two-stage reimplantation for the treatment of an infection at the site of a total knee arthroplasty, found the survival rate to be 93.5% at five years and 83.9% at ten years with removal of the implant because of infection as the end point. Wolff et al., in a report from the Mayo Clinic, reported the five-year results for twenty-one patients who had been treated for an acute infection of both knees after a bilateral total knee arthroplasty. The time between the onset of symptoms and presentation was less than three weeks for all patients. The authors found the most successful treatment to be bilateral resection arthroplasty with subsequent reimplantation. Of the eleven patients (twenty-two knees) who had been managed with débridement and retention of components, nine patients (eighteen knees) required further surgery. Meek et al.7, in a study of fifty-eight knees, evaluated patient satisfaction and functional status following the treatment of infection at the site of a total knee arthroplasty that had been performed with use of the PROSTALAC articulating spacer. After a minimum duration of follow-up of two years, only two patients had had a recurrence of infection, the mean range of motion was 87.1°, and functional scores were good.
Wear and osteolysis, especially backside wear and the possibly deleterious effects of modularity, remain a major concern related to total knee arthroplasty. The diagnosis of osteolysis about the knee can be difficult because the components can obscure the presence of the disease on plain radiographs. Nadaud et al., in a cadaveric study, found that lesions as large as 36 mm in diameter were not always discernible on standard anteroposterior and lateral radiographs but were seen on oblique radiographs. Design factors that may contribute to accelerated polyethylene wear were reported in several studies. Collier et al.8, in a study of sixty-eight primary total knee implants (AMK; DePuy) that had been retrieved at the time of autopsy or revision, found that wear was most strongly correlated with the shelf age of the polyethylene but also was associated with decreasing patient age and increasing postoperative mechanical axis varus. Noble et al., in a study of retrieved total knee implants, found that the prevalence of abrasive wear seen in association with cemented implants was higher than that seen in association with cementless implants (32% compared with 9%), suggesting that cement often contributes to damage of the tibial articulating surface. Schmalzried et al. performed a radiographic and light microscopic analysis of twenty modular, posterior stabilized polyethylene inserts and noted damage of the post in eleven cases. Puloski et al. studied nine posterior stabilized femoral components from a variety of manufacturers and found that the articulating surface of the stabilizing cam was uniformly rougher than the condylar surface (p < 0.05). Huang et al. evaluated eighty revision total knee implants (including thirty-four mobile-bearing and forty-six fixed-bearing implants) and found that the prevalence of osteolysis, generally located in the region of the posterior part of the femur, was higher in association with the mobile-bearing implants. Fehring et al. studied 2091 primary total knee replacements that had been performed with use of the Press-Fit Condylar system (Johnson and Johnson) and found five variables that were positively associated with wear and osteolysis (age, gender, polyethylene sheet processor, finishing method, and shelf age in years). The thirteen-year survival rate was 82.6%. The authors concluded that small changes in design parameters can have substantial deleterious effects. Weber et al. reviewed more than 1000 total knee replacements that had been performed with the AGC prosthesis (Biomet) (including 698 replacements that had been performed with a compression-molded monoblock design and 353 that had been performed with a ram-extruded modular design) and noted higher rates of osteolysis, radiolucent lines, and revision in association with the modular design. Modifications of the polyethylene to reduce wear and osteolysis were the topic of a few reports. Muratoglu et al. performed a surface analysis of retrieved highly cross-linked tibial inserts and found similar damage scores and significantly less elimination of machine marks in the highly cross-linked group. Colwell et al., using a knee-wear simulator, compared oxidized zirconium (OxZirc) femoral components with cobalt-chromium femoral components of the same design and found a 44% reduction in polyethylene wear in association with the OxZirc components.
With the ever-increasing number and younger age of patients undergoing total knee arthroplasty, revision total knee arthroplasty will continue to increase in frequency. In one concerning report, Sierra et al. reviewed 3251 total knee revisions that had been performed at the Mayo Clinic and found a 15.1% reoperation rate, with many patients having had multiple reoperations. The authors found this rate of reoperation to be essentially unchanged as designs and techniques evolved. Deshmukh et al. evaluated the prevalence of one of the simplest revisions, polyethylene exchange, in a study of 2000 consecutive posterior cruciate ligament-retaining total knee arthroplasties and found that tibial modularity had significant benefits during reoperation in patients with this specific knee system. Laskin reported on the use of conforming tibial components as an alternative to so-called superstabilized constrained implants and found that fifty-two of fifty-eight patients had <5° of medial-lateral instability after a minimum duration of follow-up of two years. Debate continues with regard to the superiority of stem fixation during total knee revision. Whaley et al., in a study of thirty-eight posterior stabilized revision total knee arthroplas ties that had been performed with cemented stems, reported an eleven-year component survival rate of 95.7% with revision for aseptic loosening as the end point. Shannon et al.9, in a review of sixty-three consecutive revision total knee arthroplasties that had been performed with cemented tibial and femoral components and an attached uncemented intramedullary stem, reported a 19% rate of rerevision (with six rerevisions performed for loosening, four performed for infection, and two performed for instability) after a mean duration of follow-up of six years. Four other knees had radiographic signs of loosening, for an overall mechanical failure rate of 16%. Barrack et al., in a study involving 224 solid cobaltchromium stems (in 112 patients) and sixty-two slotted titanium stems (in thirty-one patients) that had been inserted without cement during revision total knee arthroplasty, reported end-of-stem pain in association with 18.8% of the cobalt-chromium stems and 8.1% of the titanium stems (p < 0.05). Parsley et al., in a multicenter review of 199 revision total knee arthroplasties that had been performed with many different techniques of stem fixation and many different stem lengths, found that anteroposterior tibial alignment was most accurate in association with long, canal-filling stems and least accurate in association with cemented stems. Many papers and presentations dealt with the patellar component in patients undergoing revision total knee arthroplasty. Karnezis et al. compared the effect of secondary (delayed) patellar resurfacing with that of primary resurfacing at the time of the index arthroplasty and found that delayed resurfacing provided substantial clinical improvement; however, the amount of improvement was inferior to the result of primary replacement. Lonner et al.10 reviewed the results of 202 revision total knee arthroplasties in which a well-positioned, stable, allpolyethylene patellar component was retained. The authors found that retention of the patellar component can be successful, even with manufacturing mismatch, provided that the polyethylene does not show substantial wear or delamination. Meek et al. compared the results for fifty-two knees in which the patellar component was left in place following revision total knee arthroplasty with those of fifty-eight knees in which inadequate bone was present for patellar resurfacing and found no significant difference between the two cohorts with regard to pain and function scores. Leopold et al.11, in a study of forty knees that had undergone isolated patellar revision, reported a 38% rate of refailure after a mean duration of follow-up of sixty-two months. The authors reported that patellar component failure was associated with implant design and/or component alignment, both of which should be scrutinized carefully before proceeding with isolated patellar revision. In contrast, Burke et al. evaluated the outcomes of isolated revision of failed metal-backed patellar components after a minimum duration of follow-up of four years and concluded that revision of a failed metal-backed patellar component to a cemented all-polyethylene patellar component could be a durable, successful procedure. In the study by Nelson et al., sixteen patients who had been managed with a new tantalum-backed patellar component that had been developed to compensate for patellar bone loss were compared with nine patients who had been managed with patelloplasty. Favorable results with regard to diminished anterior knee pain and function were observed in the group that had received the tantalum component.
|
Top
Introduction
Treatment of Osteoarthritis...