Excessive opioid prescriptions after total joint arthroplasty (TJA) increase risks for adverse opioid related events, chronic opioid use, and increase the availability of opioids for unlawful diversion. Thus, decreasing postoperative prescriptions may improve quality after TJA. Concerns exist that a decrease in opioids prescribed may increase complications such as readmissions, emergency department (ED) visits or worsened patient reported outcomes (PROs). The purpose of this quality improvement study was to explore whether a reduction in opioids prescribed after TJA resulted in increased complications.

Methods: Data originated from a statewide arthroplasty database (MARCQI). The database collects over 96% of all TJA performed in the state of Michigan, USA. Data was prospectively abstracted and included OMEs prescribed at discharge, readmissions, ED visits within 30 days and PROs. Data was collected one year before and after the creation of an opioid prescribing protocol that had decreased prescriptions by approximately 50% in opioid naive and tolerant patients. Trends were monitored using Shewhart control charts.

84,998 TJA over two-years were included. All groups showed a reduction in opioids prescribed. Importantly, no increased complications occurred concomitant to this reduction. No increases in ED visits or readmissions, and no decreases in KOOSJR/HOOSJR/PROMIS10 scores were noted in any of the groups.

Using large data sets and registries can drive performance and improve quality. The MARCQI Postoperative opioid prescription recommendations and performance measures decreased total oral morphine equivalents prescribed over a large and diverse population by approximately 50% without decreasing PROs or increasing ED visits or hospital readmissions. A reduction in opioids prescribed after TJA can be accomplished safely and without an increase in complications across a large population.

Failure of osseointegration and periprosthetic joint infection (PJI) are the two main reasons of implant failure after total joint replacement (TJR). Nanofiber (NF) implant surface coating represents an alternative local drug eluting device that improves osseointegration and decreases the risk of PJI. The purpose of this study was to investigate the therapeutic efficacies of erythromycin (EM)-loaded coaxial PLGA/PCL-PVA NF coating in a rat S. aureus-infected tibia model.

NF coatings with 100mg and 1000mg EM were prepared. NF without EM was included as positive control. 56 Sprague Dawley rats were divided into 4 groups. A titanium pin (1.0-mm x 8 mm) was placed into the tibia through the intercondylar notch. S. aureus (SA) was introduced by both direct injection of 10 μl broth (1 × 10⁴ CFU) into the medullary cavity and single dip of Ti pins into a similar solution prior to insertion. Rats were sacrificed at 8 and 16 weeks after surgery. The outcome measurements include μCT based quantitative osteolysis evaluation and hard tissue histology.

Results: EM-NF coating (EM100 and EM1000) reduced osteolysis at 8 and 16 weeks, compared to EM0 and negative control. The effective infection control by EM-NFs was further confirmed by hard tissue section analysis. The Bone implant contact (BIC) and bone area fraction Occupancy (BAFO) within 200 µm of the surface of the pins were used to evaluate the osseointegration and new bone formation around the implants. At 16 weeks, the bone implant contact (BIC) of EM 100 (35.08%) was higher than that of negative control (3.43%) and EM0 (0%). The bone area fraction occupancy within 200 µm (BAFO) of EM100 (0.63 mm2) was higher than that of negative control (0.390 mm2) and EM0 (0.0 mm²). The BAFO of EM100 was also higher than that of EM1000 (0.3mm²).

There was much less osteolysis observed with EM100 and EM1000 NF coatings at 16 weeks, as compared to EM0 positive control, p=0.08 and p=0.1, respectively. Osseointegration and periprosthetic bone formation was enhanced by EM-NFs, especially EM100. Data from this pilot study is promising for improving implant surface fabrication strategies.

The efficacy of saline irrigation for the treatment of periprosthetic infection (PJI) is limited in the presence of infected implants. This study evaluated the efficacy of vancomycin/tobramycin-doped polyvinyl alcohol (PVA)/ceramic composites (PVA-VAN/TOB-P) after saline irrigation in a mouse pouch infection model.

3D printed porous titanium (Ti) cylinders (400, 700 and 100 µm in pore size) were implanted into mice pouches, then inoculated with S. aureus at the amounts of 1X10³ CFU and 1X10⁶ CFU per pouch, respectively. Mice were randomized into 4 groups (n=6 for each group): (1) no bacteria; (2) bacteria without saline wash; 3) saline wash only, and (4) saline wash+PVA-VAN/TOB-P. After seven days, pouches were washed out alone or with additional injection of 0.2 ml of PVA-VAN/TOB-P. Mice were sacrificed 14 days after pouch wash. Bacteria cultures of collected Ti cylinders and washout fluid and histology of pouch tissues were performed.

The low-grade infection (1X10³ CFU) was more significant in 400 µm Ti cylinders than that in Ti cylinders with larger pore sizes (700 and 1000 µm (p<0.05). A similar pattern of high-grade infection (1X10⁶ CFU) was observed (p<0.05). For the end wash, the bacteria burden (0.49±0.02) in saline wash group was completely eradicated by the addition of PVA-VAN/TOB-P (0.005±0.001, p<0.05).

We noticed that 400 µm Ti cylinders have the highest risk of implant infection. Our data supported that the effect of saline irrigation was very limited in the presence of contaminated porous Ti cylinders. PVA-VAN/TOB-P was biodegradable, biocompatible, and was effective in eradicating bacteria retention after saline irrigation in a mouse model of low grade and high-grade infection. We believe that PVA-VAN/TOB-P represents an alternative to reduce the risk of PJI by providing a sustained local delivery of antibiotics.

Irrigation with antiseptic agents, antibiotics, and surfactants are used for treatment and prevention of infections. Despite desirable microbicidal actions, studies have demonstrated cytotoxic effects on host tissue that may impair healing. This study investigated the extent of tissue damage caused by commonly used irrigation solutions in the presence or absence of infection.

Air pouches created in 60 balb/c mice were divided into two groups (n=30): infected with Staphylococcus aureus and control. One week later the infected group was subdivided into 5 subgroups (n=6) based on irrigation solutions and by day 0 (immediately) and day7 after irrigation (n=3). Solutions included Saline, Bacitracin, Clorpactin, Irrisept and Bactisure. In infected group wash fluid was collected for quantitative analysis of bacterial growth. At the specified times mice were sacrificed, pouch tissue sent for histology, and sections analyzed for inflammation, necrosis, and edema.

Inflammation decreased in infected vs sterile pouches for all solutions except Bacitracin day 0 and for all solutions day 7 with significance in all except Bacitracin (p<0.05). On day 0, necrosis increased in infected vs sterile pouches in Bacitracin (p=0.006), Irrisept (p=0.18), or Bactisure (p=0.07); however, on day 7, necrosis significantly decreased in infected pouches for all solutions (p<0.05) except for Clorpactin (p=0.18). Edema decreased in infected vs sterile pouches on day 0 for all solutions with significance in saline, Irrisept, and Bacitracin (p<0.05). On day 7, infected pouches had decreased edema in saline, Bacitracin, and Bactisure (p<0.05) and increased in Irrisept (p<0.05) and Clorpactin (p=0.069) compared to sterile pouches. Bacterial culture of washouts demonstrated that Clorpactin, Irrisept and Bactisure controlled the infection, whereas saline and Bacitracin showed bacterial multiplication 3.9 × 10^7 CFU/ml and 6.7 × 10^7 CFU/ml respectively. Bacitracin wash showed significantly more bacteria growth compared to Clorpactin (p=0.024), Irrisept (p=0.025) and Bactisure (p=0.025).

Tissue damage varied with irrigation solutions and the presence or absence of infection. Presence of bacteria appeared to lead to less tissue inflammation and edema. Tissue necrosis varied over time with different solutions. Surgeons must weigh risks and benefits when selecting solutions and determining when to irrigate.

Extensor mechanism and abductor reconstructions in total joint arthroplasty are problematic. Growing tendon into a metallic implant would have great reconstructive advantages. With the introduction of porous metal implants, it was hoped that tendons could be directly attached to implants. However, the effects of the porous metal structure on tissue growth and pore penetration is unknown. In this rat model, we investigated the effect of pore size on tendon repair fixation using printed titanium implants with differing pore sizes.

There were four groups of six Sprague Dawley rats (n = 28) plus control (n=4). Implants had pore sizes of 400µm (n=8), 700µm (n=8), and 1000µm (n=8). An Achilles tendon defect was created, and the implant positioned and sutured between the cut ends. Harvest occurred at 12-weeks. Half the specimens underwent tensile load to failure testing, the other half fixed and processed for hard tissue analysis.

Average load to failure was 72.6N for controls (SD 10.04), 29.95N for 400µm (SD 17.95), 55.08N for 700µm (SD 13.47), and 63.08N for 1000µm (SD 1.87). The load to failure was generally better in the larger pore sizes. Histological evaluation showed that there was fibrous tendon tissue within and around the implant material, with collagen fibers organized in bundles. This increases as the pore diameter increases.

Printing titanium implants allows for precise determination of pore size and structure. Our results showed that tendon repair utilizing implants with 700µm and 1000µm pores exhibited similar load to failure as controls. Using a defined pore structure at the attachment points of tendons to implants may allow predictable tendon to implant reconstruction at the time of revision arthroplasty.

Introduction

Total knee arthroplasty is very successful although the clinical assessment and rated outcome does not always match the patients reported satisfaction. One reason for patient dissatisfaction is less than desired range of motion. Poor postoperative motion inhibits many functional activities and may create a perception of dysfunction. Early in the postoperative period when patients are having trouble regaining motion (usually 6–8 weeks), manipulation under anesthesia can be used to advance range of motion by manually lysing adhesions.

Comorbidities have been used as predictors for outcome in total knee arthroplasty in population health studies. Likewise, predicting which patients are most susceptible to early postoperative stiffness/manipulation would be valuable for patient education and to predict outcome.

Bone ingrowth is desired with uncemented hip implants. Infection is clearly undesirable. We have worked on developing a nanofiber coating for implants that would enhance bone formation while inhibiting infection. Few studies have focused on developing an implant surface nanofiber (NF) coating to prevent infection and enhance osseointegration by local drug release. In this study, coaxial doxycycline (Doxy)-doped polycaprolactone/polyvinyl alcohol (PCL/PVA) Nanofibers were directly deposited on the titanium (Ti) implant surface during electrospinning.

The interaction of loaded Doxy with both PVA and PCL NFs was characterized by Raman spectroscopy. The bonding strength of Doxy-doped NF coating on Ti implants was confirmed by a stand single-pass scratch test. The improved implant osseointegration by PCL/PVA NF coatings in vivo was confirmed by scanning electron microscopy, histomorphometry and micro computed tomography at 2, 4 and 8 weeks after implantation. The bone contact surface (%) changes of NF coating group (80%) is significantly higher than that of no NF group (< 5%, p<0.05). Finally, we demonstrated that Doxy-doped NF coating effectively inhibited bacterial infection and enhanced osseointegration in an infected (Staphylococcus aureus) tibia implantation rat model. Doxy released from NF coating inhibited bacterial growth up to 8 weeks in vivo. The maximal push-in force of Doxy-NF coating (38 N) is much higher than that of NF coating group (6.5 N) 8 weeks after implantation (p<0.05), which was further confirmed by quantitative histological analysis and micro computed tomography.

These findings indicate that coaxial PCL/PVA NF coating doped with Doxy and/or other drugs have great potential in enhancing implant osseointegration and preventing infection.

Background

Total knee prostheses are continually being redesigned to improve performance, longevity and closer mimic kinematics of the native knee. Despite continued improvements, all knee implants even those with proven design features, have failures. We identified a cohort of patients with isolated tibial component failures that occurred in a popular and successful knee system. Our purpose was to (1) characterize the observed radiographic failure pattern; (2) investigate the biologic response that may contribute to the failure; and (3) to determine if the failure mechanism was of a biological or a mechanical nature.

Introduction

Alternative payment models, such as bundled payments, aim to control rising costs for total knee (TKA) and total hip arthroplasty (THA). Without risk adjustment for patients who may utilize more resources, concerns exist about patient selection and access to care. The purpose of this study was to determine whether lower socioeconomic status (SES) was associated with increased resource utilization following TKA and THA.

Few studies have been reported focusing on developing implant surface nanofiber (NF) coating to prevent infection and enhance osseointegration by local drug release. In this study, coaxial doxycycline (Doxy)-doped polycaprolactone/polyvinyl alcohol (PCL/PVA) NFs were directly deposited on the titanium (Ti) implant surface during electrospinning. The bonding strength of Doxy-doped NF coating on Ti implants was confirmed by a stand single-pass scratch test. The improved implant osseointegration by PCL/PVA NF coatings in vivo was confirmed by scanning electron microscopy, histomorphometry and micro computed tomography at 2, 4 and 8 weeks after implantation. The bone contact surface (%) changes of NF coating group (80%) is significantly higher than that of no NF group (< 5%, p<0.05). Finally, we demonstrated that Doxy-doped NF coating effectively inhibited bacterial infection and enhanced osseointegration in an infected (Staphylococcus aureus) tibia implantation rat model. Doxy released from NF coating inhibited bacterial growth up to 8 weeks in vivo. The maximal push-in force of Doxy-NF coating (38 N) is much higher than that of NF coating group (6.5 N) 8 weeks after implantation (p<0.05), which was further confirmed by quantitative histological analysis and micro computed tomography. These findings indicate that coaxial PCL/PVA NF coating doped with Doxy and/or other drugs have great potential in enhancing implant osseointegration and preventing infection.

Background

The ability to identify those at risk for longer inpatient stay helps providers with postoperative planning and patient expectations. Decreasing length of stay in the future will be determined by appropriate patient selection, risk stratification, and pre-operative patient optimization. The purpose of this study was to identify factors that place patients at risk for extended postoperative lengths of stay.

Introduction:

A pulmonary embolus (PE) occurs frequently in medical patients and acutely in post surgical total joint patients. While the two groups seem vastly different, there has never been an analysis of the location, number of emboli a patient has and size of the emboli in post-operative total joint arthroplasty patients compared to general medical patients. Studies have looked at the size of PEs relative to symptomatic patients; recent data have suggested the timeline of PE development is usually within the first few days after total knee arthroplasty (TKA) or total hip arthroplasty (THA).

We have demonstrated that erythromycin (EM) inhibits wear debris-induced macrophage activation and osteo-clastogenesis (both in vitro and in vivo) through targeting NF-κB signalling. Our clinical trial further verified that oral EM can be efficiently delivered to periprosthetic tissue and improve local inflammation. The purpose of this study was to assess the efficacy of periprosthetic EM delivery in a rat osteolysis model.

The PA coated titanium (Ti) pin (Stryker) was loaded with EM (8 μl = 2.8 mg/pin). Drug release assay showed around 25% of loaded EM was remained in the PA layer 24 hours after loading. Rats were divided into three groups:

saline control (n=5);

Uncoated Ti pins were pressfit inserted into right tibia following the injection 200 μl of either UHMWPE particles (5 mg/ml) or saline (control). The revision surgeries were performed 6 weeks after the first surgery. The previous implanted pins were replaced with new Ti pins either with or without EM coating. Rats were then sacrificed one month after “revision surgery”, and the knee joint samples were collected for μCT and histology analysis.

μCT analysis showed that the value of bone volume (bv/tv) in the group treated with EM (0.26 ± 0.07) was significantly higher than the group untreated (0.14 ± 0.04), while there was no significant difference between EM treated group and the saline control group (0.15 ± 0.11). The parameters of cancellous bone structure all pointed a trend of better structure in EM treated group than other two groups. However, this difference did not reach statistical significance. Histology analysis (H& E staining) demonstrated that in the saline control the tibia retained a smooth endocortical surface with a prominent periprosthetic membrane. In the EM-treated group, endocortical erosion was reduced and the peri-prosthetic tissue appeared thinner than uncoated pins.

The overall cellularity of periprosthetic membranes from the EM-treated group was decreased compared to the untreated group. Analysis of membrane thickness revealed a significantly thinner membrane in EM-treated group compared with untreated group and saline control (p< 0.05).

The results of this study seem to indicate that an EM coated Ti pin provided a sufficient drug source to effectively treat wear debris-induced periprosthetic inflammation and osteolysis.

Numerous investigators have described chondrogenic differentiation of bone marrow stromal cells obtained from both murine and human sources over the past decade. The ease of access and large available quantity of adipose tissue, however, makes Adipose-Derived Stem Cells (ADSC) a far more practical alternative for clinical applications.

Therefore, the primary goal of this research endeavor is to achieve chondrogenic differentiation of ADSC. Previous work had also demonstrated that bone morphogenetic protein receptor 1A (BMP receptor 1A) signaling is required for postnatal maintenance of articular cartilage. In fact, cartilage within the joints of transgenic mice deficient in BMP receptor 1A rapidly degenerates after birth in a process resembling accelerated human osteoarthritis. Based on this evidence, we used a lentiviral vector to increase expression of BMP receptor 1A by our isolated stem cells in order to direct their differentiation into the chondrocyte lineage.

We harvested subcutaneous adipose tissue intraoperatively from consenting patients undergoing elective lipoplasty and panniculectomy procedures. The stromal vascular fraction was isolated from this tissue and further refined by passaging in selective media to yield a stable population of ADSC in primary culture. Both the identity and homogeneity of this stem cell population was confirmed using adipogenic induction media and differentiation cocktails. In addition, we subcloned an expression plasmid containing the BMP receptor 1A locus in tandem with green fluorescent protein (GFP) under the transcriptional control of a single promoter. This plasmid was packaged into a lentiviral vector to provide a reliable method of achieving both genomic integration and long-term expression of the BMP receptor 1A gene. Hence, transduction of ADSC using this vector resulted in overexpression of BMP receptor 1A by these multipotent cells.

The GFP was then utilized to screen and enrich the ADSC population for stem cells with a robust expression of BMP receptor 1A. The ADSC that overexpressed BMP receptor 1A were found to achieve chondrogenic differentiation after 13 to 16 days of in vitro culture, as revealed by immunohistochemistry assays for the bio-markers of articular cartilage (type II collagen and the proteoglycan aggrecan).

Our results demonstrate that stem cells derived from the adipose tissue of a patient represent a viable means of culturing autologous chondrocytes in vitro for future implantation at the site of osteochondral defects. This method of attaining cartilaginous regeneration is intuitively appealing, given the minimal donor site morbidity associated with removing subcutaneous fat. By transducing the ADSC with a lentiviral vector, we have also collected further evidence implicating the critical importance during chondrogenesis of signaling mediated by the BMP receptor 1A. Further tissue engineering studies are now in progress to evaluate the ability of ADSC to differentiate into chondrocytes after seeding onto poly-caprolactone polymer scaffolds.

Neutron beam irradiation is currently being explored as an alternative modality to improve local control of sarcomas. The purpose of this study was to investigate the effects of a sarcoma-dose fast neutron therapy on the wear properties of standard and highly cross-linked polyethylene total hip arthroplasty liners

Two groups of 28 mm I.D. polyethylene liners were used in this study – conventional polyethylene liners (N2vac: 3Mrads innitrogen, Howmedica Osteonic, Allendale, NJ), and highly cross-linked liners (Crossfire: 10.5MRads total radiation dose, Howmedica Osteonics, Allendale, NJ). All liners were sterilized in a oxygen free environment and stored in inert nitrogen packages. The plastic cups were sandwiched between two tissue-equivalent blocks to simulate the human hip region and brought to the fast neutron therapy unit. The neutron beam is produced in a super conducting cyclotron by bombarding an internal beryllium target with 48.5 MeV deuterons [d(48.5)+BE]. The cups were exposed to a dose of 15 Gy represented a typical neutron-dose given to a sarcoma patient. Wear testing was then performed utilizing a hip simulator (MTS, EdenPrairie, MN) with matched 28 mm diameter CoCr femoral heads. Physiologic loading was simulated with biaxial cross-path motion and peak loads of 2450 N. All tests were performed in 50 percent diluted alpha-calf serum(Hyclone Laboratories, Logan, UT) to simulate human serum exposure. Every 250,000 cycles the serum was changed and samples were removed from the machine, cleaned and weighed. The volume loss measurement shown below used the weight loss to calculate the wear rate. The wear rate was converted to volume loss by dividing by the density. The value is given as millimeter scubed per million cycles (mm3/mc). Phase one cups were tested within one month of radiation. Phase two cups served as soaked controls, and spent 7months in calf serum prior to wear testing. A total of five million wear cycles were performed for each cup to simulate five years worth of use.

The Averaged volumetric wear loss data demonstrated significantly less wear in CrossfireÒ compared to N2vac in both neutron irradiated and non-irradiated samples. This suggests that in sarcoma cases of the hip involving adjuvant fast neutron therapy, highly crosslinked poly-ethylene should be utilized. Averaging all data there was no statistically significant difference between the neutron radiation and non-treated components for both material conditions (N2VacÒ and CrossfireÒ). A trend towards decreasing wear in phase two samples was noted which may represent a material change in the liners exposed to serum over time. In addition, larger than normal variability in wear rates was seen within each group. Further testing of these liners is planned to elucidate these phenomenon. Table 1: – Volumetric wear loss per group. UHMWPE Material Volumetric wear Loss (mm3/mc) Standard Deviation N2VacÒ Phase 1 29.6 1.6 N2VacÒ Phase 2 14.0 N/a Neutron treated N2VacÒ Phase 1 52.2 18.1 Neutron treated N2VacÒ Phase 2 20.9 0.8 CrossfireÒ Phase 1 3.0 0.9 CrossfireÒ Phase 2 2.0 0.6 Neutron treated CrossfireÒ Phase 1 2.5 0.5 Neutron treated CrossfireÒ Phase 2 1.9 0.03 Graph 1– Averaged volumetric wear loss values

Sarcoma-dose (15Gy) fast neutron therapy adversely affects the wear of standard polyethylene acetabular cups. The wear rate of the conventional gamma-inert sterilized polyethylene increased by more than 50% following a 15Gy fast neutron treatment. The highly cross-linked polyethylene (Crossfire), on the other hand, was immune to the effect of neutron treatment at the same dose. The mechanisms responsible forth is difference are unclear and warrant further investigation. The clinical implication of this study is that for sarcoma patient receiving total hip replacement, highly crosslinked poly-ethylene rather than standard polyethylene should be used for the acetabular cup.

Introduction A biomechanical model was developed to measure wear of all-polyethylene patellar components as it relates to femoral component mal-rotation. The model, based on high load and flexion activities such as stair climbing, was used to differentiate the effects of femoral mal-rotation and differing materials on a single patellar design.

Methods The patellar components (Scorpio®, Stryker-Howmedica-Osteonics) were cemented onto metal fixtures and articulated against “aligned” and “mal-aligned” (six degrees internally rotated) femoral components. The patellar components were subjected to a constant force and articulated against femoral components flexing from 600 to 1200. Patellae of identical geometry, made of conventional and highly cross-linked ultra-high molecular weight polyethylene, were tested to 1x106 cycles. Following testing, patellar wear was determined by gravimetric measurement relative to soaked control specimens.

Results All conventional polyethylene patellae demonstrated damage in the form of burnishing and scratching of the articular surface. The mal-aligned conventional ultra-high molecular weight patellae demonstrated increased weight loss or wear relative to the aligned components (p=.048). All rotationally mal-aligned highly cross-linked polyethylene components sustained polyethylene fracture or catastrophic failure of the cement-polyethylene construct.

Conclusions Rotational mal-alignment of the femoral component will result in increased wear of polyethylene patellar components. The newer highly cross-linked materials failed to resolve this wear problem and sustained catastrophic failure when mal-aligned. Attention needs to be given to the patella-femoral articulation when implanting knee components and when developing new polyethylene as the forces in this articulation may result in polyethylene behaviour that varies dramatically from the femoraltibial articulation.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction Minimally invasive total knee arthroplasty has gained interest among total joint surgeons across the USA. It remains, however, somewhat unclear how to define minimally invasive. Small incisions may be a focal point of the surgery, but unless one is performing “limited incision surgery”, the skin does not define minimalness. The techniques and instrumentation for performing a minimally invasive total knee arthroplasty are presented.

Methods The principles of a minimally invasive total knee arthroplasty procedure include: a small incision, a vastus-medialis split, mobilization of the patella without eversion, and the use of modified instruments that have been designed for use with standard implants. The implantation instruments are downsized versions of a standard system and are easily applied to the routine arthroplasty patient.

Results The mid-vastus split appears to protect the quadriceps’ function and promote early recovery relative to the more traditional para-patellar arthrotomy. By avoiding patellar eversion altogether or only everting for a limited period, in extension, at the end of the case, elongation of the quadriceps fibers is avoided. These two measures appear protective and may promote early muscle recruitment during the post-operative recovery phase. The miniaturization of the familiar easier. Early review of post-operative radiographs has not revealed detriment to implant positioning.

Conclusions While only developmental, this technique appears amenable to navigation systems that may further the minimalization by avoiding violation of the femoral and tibial canals. Minimally invasive total knee arthroplasty needs to be defined as a concept and as to what role it should play in a surgeon’s armamentarium. While in its infancy as a technique, the new instruments and refined surgical approach appear to allow rapid rehabilitation and patient satisfaction without sacrificing positioning or outcome.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction Unicompartmental knee arthroplasty has become increasingly popular in the USA. Minimally invasive techniques for implant placement has augmented the procedure and allowed for rapid rehabilitation and return to activities of daily living. Nevertheless, with new technologies come learning curves and rediscovery of the past. Complications do occur with placement of unicompartmental knee devices. Examples of proper and improper implantation techniques and the radiographic results are presented.

Methods I present the results of a review of the literature and personal experience.

Results Overcorrection deformity has been identified as a precursor to early failure. When sizing and placing the implants an attempt should be made to “leave alignment alone” and there should be relative pseudo-laxity of the joint with the new implants. Problems with varus or valgus tibial cuts are well known. Posterior slope is less understood. While it is reasonable to reproduce the natural inclination of the tibia, over-correction or under-correction leads to balance abnormalities. Improper slope may be a prelude to subsidence particularly if an inset design is employed. Coalescence of the pin tracts used to fix cutting blocks and sagittal tibial cuts (along the spine) have been identified as problematic. The small surface area and stress loads on the tibia predispose this area to fracture, particularly when these stress risers are present. Patellar impingement can lead to pain and disability. It may be avoided by appropriate sizing, slight recession of the femoral component, or a modest resection of the medial facet. The posterior cruciate ligament is at risk during resection of the proximal tibia. Injury to the ligamentous complex will lead to instability problems not manifest in the more conforming articulation of a TKA. One unavoidable problem is the relatively large tibial resection required in “small” knees. While we attempt to be minimal in the tibial bone resections, six to eight millimetres appears large in a diminutive knee. The most difficult positioning problem appears to be internal-external rotation positioning. Several implant systems utilize the tibial cut to position the femur. The tibial platform matches or links the femoral varus-valgus (correct or not) to the tibial cut. The rotation is more free-hand and is not well coordinated by the landmarks used in TKA. Edge loading will result from rotational malposition.

Conclusions The techniques for placement, the instrumentation, and the unicompartmental implant designs have evolved to the point where many of the problems encountered in the USA in the 1980’s have been alleviated. Attention to common positional and implantation errors will result in more satisfactory outcomes. While the less conforming articulation of these devices is forgiving, it may also penalize in the long term. Many of the problems encountered after unicompartmental knee arthroplasty could be avoided with the simple awareness that promotes improved surgical technique.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Rotational mal-alignment of the patella-femoral interface will result in increased wear. Highly cross-linked polyethylene will decrease wear even if mal-aligned.

A biomechanical model based on high load and flexion was used to measure wear of rotationally aligned and mal-aligned all-polyethylene patellae. The components were articulated against “aligned” and “mal-aligned” (60 internally rotated) femoral components. The patella were subjected to a constant 2224 N force and the femoral components rotated from 600 to 1200 at 1.33 Hz. Patellae of identical geometry made of conventional UHMWPE and highly crosslinked UHMWPE were tested to 1 000 000 cycles. Wear was determined by gravimetric measurement relative to cemented soak controls.

Conventional UHMWPE: All samples demonstrated damage (burnishing and scratching) of the articulating surfaces. There was a significant increase in wear (p< .05) in the mal-aligned patella.

Highly cross-linked patellae: All components fractured in the mal-aligned construct (gamma irradiated remelted n=6, gamma irradiated and annealed n=2). Failure first occurred at the cement interface then at the posts.

Correct femoral rotation is important during TKA. The intertrochlear line, tibial cut, epicondylar axis and posterior condyles are helpful landmarks, but there is still eyeball control of rotation. It is clear from this study that rotational mal-alignment will result in increased polyethylene wear.

Highly cross-linked polyethylene has decreased wear in THA. Unfortunately, the decrease in ductility and toughness may make the use of these materials unsuitable for TKA. Based on this study model, patellar components would need to be redesigned if highly cross-linked polyethylene were to be applied.

The wear rate of conventional UHMWPE patellae is increased by rotational mal-alignment. Highly cross-linked components were a poor solution to problem. Use of highly cross-linked polyethylene resulted in component fracture.

Initial fixation affects ingrowth of uncemented tibial components. Previous studies assessed initial fixation, however most used limited one-dimensional motion measurements. Therefore, a three-dimensional micro-motion and migration analysis was performed to compare initial fixation of four different tibial tray configurations: 1) Keeled component 2) Non-keeled component (post but no keel) 3) Keeled component with screws 4) Non-keeled component with screws.

Osteonics series 7 000 tibial trays (identical to Scorpio design) with and without standard keels were obtained. The 30 mm posts were left intact. The components were implanted without cement into twenty-four fresh-frozen cadaveric tibiae. Specimens were loaded through matched femoral components via two separate loading conditions: 1) sinusoidal medial load from 200 to 2 200 N at 0.5 Hz and 2) sinusoidal torsion load from −5 Nm to +5 Nm at 0.5 Hz with constant 1 200 N axial load. Motion data were collected from reference cubes and transducers rigidly attached to the trays, during 3 000 cycles of loading. Rigid-body mechanics were used to calculate the motion of a wire-frame computer model. Comparisons were made of micro-motion and migration magnitudes.

A keel did not significantly decrease micromotion or migration under medial or torsion loading compared to a post alone (p=1.04). The addition of four cancellous screws decreased motion (p< .05) regardless of the presence of a keel under medial loading. No significant differences were noted under torsional load for any tray configuration.

A keel did not enhance initial fixation compared to screws or a post alone. Screws enhanced fixation under axial offset loading in all constructs. Screw fixation did not affect motion under torsion loads perhaps due to the limited magnitude of torque transmitted across the unconstrained femoral-tibial articulation. Based on these data, initial fixation of uncemented tibial implants should include screw fixation to counter the effects of offset axial loads.