Introduction: The increasing numbers and costs of lower limb arthroplasty procedures has lead to medical and administrative pressures to reduce lengths of stay (LOS) and the hospital costs associated with these procedures. We compared a prospective application of an arthroplasty care pathway in a tertiary referral unit in the United States (US) to a retrospective review of standard practice in an Irish unit to assess reasons for delayed discharge.

Methods: We performed the study in 2 orthopaedic units, one in the US and one in Ireland. In the US an arthroplasty perioperative care pathway was implemented, which among other elements involved a coordinated undertaking by therapists, surgeons and social workers to achieve discharge on the 3rd postoperative day. A consecutive series of primary total knee and hip arthroplasty patients were entered into that limb of the study and had demographic, clinical and LOS data collected. We then collected the same data from a randomly selected consecutive series of primary total knee and hip arthroplasty patients in the Irish unit.

Results: In the US orthopaedic unit 43 patients were included, 24 women and 19 men, of average age 60.5 years (range, 36–82). The average LOS was 3.6 days (range, 2–10), with 27 patients going to a rehabilitation facility and 16 directly home.

In the Irish orthopaedic unit 42 consecutive hip and knee arthroplasty patients from a single consultant were included. There were 26 women and 16 men with average age of 63.9 years (range 36 to 88 years). The average LOS was 6.5 days (range, 3 to 10 days), with 24 patients going to a rehabilitation facility and 18 directly home.

There was no correlation found between LOS and either comorbidities, social factors or complications, in both groups although one patient had a delayed discharge due to haematoma and wound drainage in the US. Prolonged LOS in both groups was correlated with delays in rehabilitation bed and transportation availability, reported short staffing in hospital and weekend stays.

Both groups were well matched for comparison. The average shorter LOS noted in the US unit appears to be almost entirely attributable to an implemented perioperative care pathway and a more proactive coordinated approach to discharge planning.

Discussion and Conclusion: This study demonstrates the effectiveness of multidisciplinary clinical pathways in achieving desired LOS objectives but that substantial impediments still exist from an administrative and logistical viewpoint. Such information is important for clinicians in negotiating and establishing appropriate management of their arthroplasty patients in the current care provision climate.

As the numbers of revision total knee arthroplasty (RTKA) rise, we continually need current information regarding the etiology/modes of failure and functional disability of patients presenting for RTKA. We used a prospective cohort study to assess these fundamental aspects of RTKA.

290 consecutive subjects presenting for RTKA had relevant clinical information, including modes of failure, collected from surgeon-completed documents. Patients themselves also completed quality of life and functional questionnaires, including the SF-36 and WOMAC Osteoarthritis Index.

Mean patient age was 68.6 years (55 to 79 years). Mean SF-36 and WOMAC score at baseline indicated significant functional disability. The mean time from primary procedure to RTKA was 7.9 years (6 months to 27 years). Our series included 31 percent ‘early’ (under 2 years) revisions and 69 percent ‘late’ revisions. Sepsis was the cause of 10.4 percent revisions. The tibia needed revision in 78 percent, femur in 71 percent and patella in 31 percent of cases. The predominant modes of failure (non-exclusive frequency values as patients could have more than one cause) were (in percentages): instability (28.9), malalignment (27.5), tibial osteolysis (27.5), polyethylene wear (24.5), femoral osteolysis (22.5) and tibial loosening (22.2).

These patients are relatively young, and considerably disabled by their failed primary procedure. Many modes of failure are within surgical control and direct us toward improved techniques and approaches. Other modes confirm the need for continued development of implants and materials. Information gained here will allow better formulation of measures and resource allocation that may prevent RTKAs.

The requirement for release of collateral ligaments to achieve a stable, balanced total knee replacement has been reported to arise in about 50% to 100% of procedures. This wide range reflects a lack of standardised quantitative indicators to determine the necessity for a release. Using recent advances in computerised navigation, we describe two navigational predictors which provide quantitative measures that can be used to identify the need for release. The first was the ability to restore the mechanical axis before any bone resection was performed and the second was the discrepancy in the measured medial and lateral joint spaces after the tibial osteotomy, but before any femoral resection.

These predictors showed a significant association with the need for collateral ligament release (p < 0.001). The first predictor using the knee stress test in extension showed a sensitivity of 100% and a specificity of 98% and the second, the difference between medial and lateral gaps in millimetres, a sensitivity of 83% and a specificity of 95%. The use of the two navigational predictors meant that only ten of the 93 patients required collateral ligament release to achieve a stable, neutral knee.

Introduction: A critical technical and economic challenge in total knee arthroplasty revision (TKAR) is bone loss management. Easily applied, valid pre-operative measurements of bone loss are essential to allow accurate planning and meaningful comparisons between series. We compared 2 radiographic measurement ment systems with actual intra-operative bone loss in order to determine their validity.

Methods: A prospective IRB approved cohort study of 290 consecutive TKAR patients was utilized to assess the Anderson Orthopaedic Research Institute (AORI) system and the University of Pennsylvania (UPenn) assessment system. These preoperative measures were performed on standardized antero-posterior and lateral knee radiographs by 31 orthopaedic surgeons trained in their use. The validity and reproducibility of both systems was determined versus the gold standard measure of actual intra-operative bone loss.

Results: 215 patients (74.1 percent) were assessed to have bone loss pre-operatively versus 222 (76.4 percent) intra-operatively. Using the AORI classification system agreement between preoperative and intraoperative classification was moderate for the femur (K = 0.50) and good for the tibia (K = 0.63). The UPenn system gave resultant mean scores of 0.137 for femur and tibia versus intraoperative findings of resultant mean scores of 0.14 and 0.143 for the femur and tibia, respectively. These differences were not statistically significant (p < 0.02).

Discussion: This study demonstrates a high incidence of bone loss among TKAR patients, emphasizing the importance of effective measurement tools. Both the AORI and UPenn systems are valid tools for pre-operative estimation of actual bone loss facilitating planning and clinically successful, cost effective management of bone loss in TKAR.

This study (n=126, mean age=68.8 years, males=62) evaluated pre-operative WOMAC pain and physical function, age, gender, general health status, revision severity classification, number of revisions, comorbidity and unilateral vs. bilateral surgery as predictors of WOMAC pain and physical function at twenty-four months post revision hip arthroplasty. Pain improved from 9.3 to 3.6 and physical function improved from 35.4 to 17.1. No factors were predictive of patient function. Decreased pain was predicted by less pain pre surgery (p=0.01) and being male (p=0.04).

To determine if pre-operative WOMAC pain and physical function, age, gender, general health status (SF-36), revision severity classification, number of revisions, comorbidity and unilateral vs. bilateral surgery are predictive of WOMAC pain and physical function at twenty-four months post revision hip arthroplasty.

Physical function at twenty-four months is not independently predicted by the pre-treatment factors evaluated in this study. Male patients with less pain pre surgery and little comorbidity have less pain post surgery.

With the exception of pre-treatment pain, the pre-treatment factors tested in this study provide minimal guidance in identifying factors that might be modified to enhance patient outcome.

This prospective cohort study included one hundred and twenty-six patients (mean age=68.8 years, males: females=62:64) who had revision for other than infection or peri-prosthetic fracture. On average from pre-surgery to twenty-four months post-surgery, WOMAC pain improved 9.3 to 3.6 and physical function improved from 35.4 to 17.1. In univariate analysis (t-test, p< 0.05), males tended to have better function (19.6 vs. 14.7) and reported less pain (4.4 vs. 2.8). No other factors were significant in univariate analysis. None of the a priori factors noted above were independently predictive of patient function at twenty-four months in the multivariate model (F=2.06, p=0.04, R²=0.16). Decreased pain with activity at twenty-four months independently was predicted by having less pain pre surgery (p=0.01), being male (p=0.04) and having fewer comorbidities (p=0.07) in the multi-variate model (F=2.9. p=0.004, R²=0.21).

Funding: This work was supported by a grant from The Arthritis Society

Introduction: The authors systematically reviewed the available literature in order to define the outcomes for avascular necrosis (AVN) and spontaneous osteonecrosis of the knee (SPONK) after unicompartmental knee arthroplasty (UKA) or total knee arthroplasty (TKA).

Materials and Methods: A literature review yielded seven reports with Hospital for Special Surgery (HSS) or Knee Society Score (KSS) outcomes for arthroplasty secondary to either AVN or SPONK. The mean pre-operative, post-operative, and difference in KSS or HSS scores plus the mean revision rates for the arthroplasties for each underlying disease (AVN and. SPONK) were tabulated and reported in this order. The reported means were weighted by the number of knees in each study.

Results: A total of 63 TKAs were performed for AVN and 85 TKAs were performed for SPONK. Additionally, 74 UKAs were performed for SPONK. TKAs performed secondary to AVN had mean KSS scores of 50.6, 90.2, and 39.4 points. The revision rate was 12.5% (SD=10.45). TKAs performed for SPONK had mean HSS scores of 55.6, 82.5, and 27 points. The revision rate was 5.9% (SD=2.79). UKAs performed for SPONK had mean HSS scores of 54, 83.1, and 29.1 points with a revision rate was 9.7% (SD=5.9).

Discussion: Although the KSS for TKAs performed for AVN match the KSS performed in osteoarthritic patient populations receiving TKAs, the revision rate is higher in the AVN group. The HSS scores for patients with SPONK receiving TKAs or UKAs are similar although the revision rate is higher for UKAs.

Introduction: Alcohol can induce osteoporosis and osteonecrosis. Studies have demonstrated that alcohol contributed to abnormal lipid metabolism in cells in bone marrow but the mechanisms have not been defined. The purpose of this study was to evaluate the effect of alcohol on the differentiation of pluripotential cells cloned from bone marrow.

Materials and Methods: The cells were maintained in culture and treated with either increasing concentrations of ethanol (0.09, 0.15, and 0.21 mol/L) or without alcohol to serve as controls. Morphologic features of the cells were monitored using a phase-contrast microscope. Alkaline phosphatase activity was determined using a colorimetric assay. Gene expression of adipogenesis [422 (aP2), PPAR y] and osteogenesis (osteocalcin) was evaluated using the Northern blot technique and reverse transcription-polymerase chain reaction (RT-PCR). ANOVA was used for statistical analysis.

Results: The cells treated with ethanol started to accumulate triglyceride vesicles at Day 7; the number of adipocytes and the percentage of the area that contained the cells with fat vesicles increased significantly; and the level of alkaline phosphatase activity diminished with longer durations of exposure and with higher concentrations of ethanol. Analysis of gene expression showed diminished expression of osteocalcin without a significant increase in the expression of the fat cell specific gene, 422 (aP2), and PPAR y, in cells treated with ethanol. This suggested that adipogenesis may occur at a point downstream in the fatty acid metabolism pathway.

Discussion: Alcohol induces bone marrow fatty changes in patients and in animal models contributing to osteoporosis and osteonecrosis. This study demonstrated that alcohol treatment decreased osteogenesis while enhancing adipogenesis by bone marrow stromal cells, which may be one of the mechanisms leading to osteoporosis and osteonecrosis. Inhibition of adipogenesis may lead to the prevention of the disease.

Clinical relevance: This is a novel finding that alcohol induces adipogenesis in a cloned bone marrow stromal cell. The results explain the clinical observation that there is increased adipogenesis in alcohol-induced osteoporosis and osteonecrosis.

Introduction: Treatment of osteonecrosis continues to be a challenging problem in orthopedic practice. Arthroplasty is generally successful but long-term results are inferior especially in young adults. Alternative treatments such as core decompression and trap-door procedures provide only temporary benefits and need much improvement. The replacement of necrotic bone to promote osteogenesis and angiogenesis and healing subchondral bone are future approaches. Autogenous cancellous bone is the preferred graft material but its supply is limited. Allografts are useful but not as desirable as autografts. Substitutes for bone grafts have been actively researched but few are available currently. In this study, we have attempted to use genetically engineered bone marrow stem cells in order to enhance the healing of a bone defect in a mouse model.

Materials and Methods: A bone marrow stem cell was cloned from Balb/c mice and transfected with LacZ and neomycin resistance genes. The cells were cultured for 7 to 10 days and both the osteoblastic and angiogenic properties of the cells were examined using Northern blots to detect osteocalcin and VEGF gene expression. The cells were also analyzed for alkaline phosphatase activity to demonstrate the osteoblastic phenotype of the cells. A suspension containing 2 x 10⁷ cells/ml phosphate buffered solution was prepared for cell transplantation. A total of forty-eight, 8-week old Balb/c mice were used in this study. A 1.2 mm defect was created bilaterally with an electric drill in the femurs of 24 mice to mimic the core decompression and trap-door procedures. 2 x 10⁶ cells were transplanted into each defect of the right femur while the left femur served as a control trap-door defect which was injected with PBS but without cells. An equal number of cells were injected either at subcutaneous sites, in the hindquarter muscles, or into the renal capsule (8 mice in each site) to evaluate ossification at ectopic sites. Animals were sacrificed at 2, 4, 6 and 8 weeks. Defect repair was evaluated radiographically and the contribution to osteogenesis by transplanted cells was studied histomorphometrically using tissue sections stained with X-gal as well as biochemically on DNA extracts using primers for the neomycin resistance gene.

Results: Radiopaque tissue appeared two weeks after the cells were transplanted into bone defects, muscle, subcutaneous sites, and the renal capsule. Histological analysis demonstrated that these tissues consist of newly formed bone from transplanted cells that stained positively with X-gal and contained neo DNA. The repair tissue did not contain cartilaginous areas indicating that ossification surrounding the D1-BAG cells was not through the endochondral process. At four weeks, 4 of 6 femora showed a defect that was filled with new bone. At 6 weeks, all of the defects (6 of 6) contained fully restored bone. However, in the control side that was injected with PBS (no cells) only 2 of 6 at 4 weeks, 3 of 6 at 6 weeks, and 5 of 6 at 8 weeks showed complete repair. All histological sections of bone defects (n = 24) were examined histomorphometrically using a computerized image analysis system. Transplantation of marrow stem cells into bone defects produced more bone at an earlier time point than controls and, the process of enhanced ossification continued throughout the healing process.

Discussion: The cloned bone marrow stem cell can directly form bone after transplantation into bone defects and into ectopic sites, indicating that the in vitro expanded bone marrow stem cells can serve as a grafting material to enhance healing of bone defects and the treatment of osteonecrosis. In addition, this study demonstrates that genetic labelling is a useful tool in studies of cell differentiation in vivo and that bone marrow stem cells may be useful as a carrier of genetically-engineered factors in the treatment of skeletal diseases.

Introduction: Osteonecrosis continues to be a challenging problem in orthopaedic practice. Etiology is multi-factorial but steroid- and alcohol-associated osteonecrosis contributed to more than two thirds of all the cases. While the pathogenesis of the disease is still unknown, many new insights have emerged from research in the last decade. Studies have demonstrated that both steroids and alcohol promote adipogenesis and inhibit osteogenesis, in vitro and in vivo, leading to osteonecrosis and osteoporosis. It has been found that Dexamethasone can turn on adipogenic transcription factor PPARy2 but suppress osteogenic transcription factor Cbfa1/Runx2. Steroids also decrease VEGF production resulting in inhibition of angiogenesis by osteoprogenitor cells. However, alcohol produces adipogenesis through a different mechanism at a point downstream in the fatty acid metabolism pathway, but it does inhibit osteogenesis by decreasing osteocalcin gene expression. Increased adipogenesis and osteoporosis, together with decreased osteogenesis and angiogenesis, will eventually lead to the final pathway of osteonecrosis.

Although conventional thinking and teaching have implicated weight and body mass index (BMI) in premature failure of total knee arthroplasty (TKA) there is scant evidence based confirmation of this belief. Furthermore, there is little knowledge regarding the precise effect of BMI on functional outcomes following TKA. We performed this study to assess the effect of weight on the longevity of TKA and on outcomes following TKA revision (TKAR).

186 consecutive subjects undergoing TKAR in a 17-center prospective cohort study, had data collected on weight (pounds), BMI and time elapsed between primary and revision surgery (T). The Physical Component Score (PCS) of the Short Form-36 (SF-36), the Western Ontario and McMaster Universities Osteoarthritis (WOMAC) Index, and the Knee Society Score (KSS) were also collected preoperatively and at 6-month follow-up. Univariate, bivariate and multivariate statistical methods were used in the analysis.

The mean BMI and weight were 31.8 (54% of subjects had a BMI > 30) and 200 pounds (range 107–350) respectively. The distribution of both measures of excessive weight was close to normal. Average time between primary and revision procedures (T) was 7.3 years (range 6 months to 27 years). Using linear regression, T significantly decreased as weight (BMI) increased. Mean SF-36 PCS, WOMAC and KSS-Function scores were significantly improved 6 months after revision surgery. However, BMI and, in particular, weight were predictive of worse physical functional outcomes.

This study demonstrates the deleterious effect of weight on both the longevity of primary TKA as assessed at the time of revision and on functional outcomes following TKAR. Although further prospective data regarding this population is indicated, the current findings direct us towards better outcomes prediction for overweight patients and more effective counselling and appropriate management of these patients.

Introduction: Whether or not to surgically treat osteonecrosis of femoral head (ONFH) when patients are asymptomatic is controversial. The goal of this study was to determine: 1) if spontaneous resolution of ONFH does occur, 2) how long does it take for resolution to occur, and 3) if there are predictors of spontaneous resolution.

Materials and Methods: For this prospective study, patients with asymptomatic ONFH were identified from two National Institute of Health funded, Institutional Review Board approved screening studies. A prospective screening study for ONFH after organ transplantation was begun in 1997 by performing routine MRI examinations after transplantation. In a second prospective study on surgical treatment for symptomatic ONFH, the contralateral hip was screened for asymptomatic disease. A cohort of patients having hips with asymptomatic ONFH was then analyzed.

Results: As of December 2000, 13 asymptomatic hips in 10 patients were identified from the prospective screening study after organ transplantation and 17 hips in 17 patients were identified from the contralateral hip screening study. There were 3 hips with ARCO stage I disease showing evidence of spontaneous resolution. The modified index of necrotic extent measured 11.10, 12.72, and 20.83, with the estimated femoral head involvement being 15–30% in 2 of the hips and less than 15% in the third. Resolution on MRI was complete in 2 of the 3 hips, and nearly complete in the third.

Discussion: Spontaneous resolution of ONFH does occur. Factors associated with resolution are early, asymptomatic disease (ARCO stage I), small lesion size (modified index of necrotic extent < 25), and the absence of symptomatic disease in the contralateral hip. Initial signs of resolution may take up to one year to occur. For patients fitting these criteria, we recommend withholding surgery and monitoring hips with serial MRI observation to monitor the course of their disease.