Introduction. Pain related to knee osteoarthritis (OA) is a complex phenomenon that cannot be fully explained by radiographic disease severity. We hypothesized that pain phenotypes are likely to be derived from a confluence of factors across multiple domains: knee OA pathology, psychology, and neurophysiological pain processing. The purpose of this study was to identify distinct phenotypes of knee OA, using measures from the proposed domains. Methods. Data from 3494 subjects participating in the Osteoarthritis Initiative (OAI) study was analyzed. Variables analyzed included: radiographic OA severity (Kellgren-Lawrence grade), isometric quadriceps strength, Body Mass Index (BMI), comorbidities, CES-D Depression subscale score, Coping Strategies Questionnaire Catastrophizing subscale score, number of pain sites, and knee tenderness on physical examination. Variables used for comparison across classes included pain severity, WOMAC disability score, sex and age. Latent Class Analysis was performed. Model solutions were evaluated using the Bayesian Information Criterion. One-way ANOVAs and post hoc least significance difference tests were used for comparison of classes. Results. A four-class model was identified. Class 1 (57% of study population) had lesser radiographic OA, little psychological involvement, greater strength, and less pain sensitivity. Class 2 (28%) had higher rates of knee joint tenderness. Class 3 (10%) had greater psychological distress and more bodily pain sites. Class 4 (4%) had more comorbidities. Additionally, Class 1 was the youngest, had the lowest disability, and least pain. Class 4 was the oldest. Class 2 had a higher proportion of females. Class 3 had the worst disability and most pain. Conclusions. Four distinct pain phenotypes for knee OA were identified. Psychological factors, knee tenderness, and comorbidities appear to be important in defining phenotypes of OA-related pain. Therapies in knee OA should take a multicomponent approach, recognizing the factors most relevant to an individual's experience of pain

Purpose. It is well known that meniscus extrusion is associated with structural progression of knee OA. However, it is unknown whether medial meniscus extrusion promotes cartilage loss in specific femorotibial subregions, or whether it is associated with a increase in cartilage thickness loss throughout the entire femorotibial compartment. We applied quantitative MRI-based measurements of subregional cartilage thickness (change) and meniscus position, to address the above question in knees with and without radiographic joint space narrowing (JSN). Methods. 60 participants with unilateral medial OARSI JSN grade 1–3, and contralateral knee OARSI JSN grade 0 were drawn from the Osteoarthritis Initiative. Manual segmentation of the medial tibial and weight-bearing medial femoral cartilage was performed, using baseline and 1-year follow-up sagittal double echo steady-state (DESS) MRI, and proprietary software (Chondrometrics GmbH, Ainring, Germany). Segmentation of the entire medial meniscus was performed with the same software, using baseline coronal DESS images. Longitudinal cartilage loss was computed for 5 tibial (central, external, internal, anterior, posterior) and 3 femoral (central, external, internal) subregions. Meniscus position was determined as the % area of the entire meniscus extruding the tibial plateau medially and the distance between the external meniscus border and the tibial cartilage in an image located 4mm posterior to the central image (a location commonly used for semi-quantitative meniscus scoring). The relationship between meniscus position and cartilage loss was assessed using Pearson (r) correlation coefficients, for knees with JSN and without JSN. Results. The percentage of knees showing a quantitative value of >3mm medial meniscus extrusion was 50% in JSN knees, and only 12% in noJSN knees. The 1-year cartilage loss in the medial femorotibial compartment was 74±182µm (2.0%) in JSN knees, and 26±120µm (0.8%) in noJSN knees. There was a significant correlation between cartilage loss throughout the entire femorotibial compartment (MFTC) and extrusion area in JSN knees but not for noJSN knees. Also, the extrusion distance measured 4mm posterior to the central slice was not significantly correlated with MFTC cartilage loss. The strongest (negative) correlation between meniscus position and subregional femorotibial cartilage loss (r=−0.36) was observed for the external medial tibia. In contrast, no significant relationship was seen in the central tibia. No significant relationship was found in other tibial subregions, except for the anterior medial tibia, but only in JSN knees (r=−0.27). Correlation coefficients for the femoral subregions were generally smaller than those for tibial subregions, with only the internal medial weight-bearing femur attaining statistical significance (r =−0.26). Conclusions. The current results show that the relationship between meniscus extrusion and cartilage loss differs substantially between femorotibial subregions. The correlation was strongest for the external medial tibia, a region that is physiologically covered by the medial meniscus. It was less for other tibial and femoral subregions, including the central medial tibia, a region that exhibited similar rates of cartilage loss as the external subregion. The findings suggest that external tibia may be particularly vulnerable to cartilage tissue loss once the meniscus extrudes and the surface is “exposed” to direct, non-physiological, cartilage-cartilage contact

Background. Artificial total knee designs have revolutionized over time, yet 20% of the population still report dissatisfaction. The standard implants fail to replicate native knee kinematic functionality due to mismatch of condylar surfaces and non-anatomically placed implantation. (Daggett et al 2016; Saigo et al 2017). It is essential that the implant surface matches the native knee to prevent Instability and soft tissue impingement. Our goal is to use computational modeling to determine the ideal shapes and orientations of anatomically-shaped components and test the accuracy of fit of component surfaces. Methods. One hundred MRI scans of knees with early osteoarthritis were obtained from the NIH Osteoarthritis Initiative, converted into 3D meshes, and aligned via an anatomic coordinate system algorithm. Geomagic Design X software was used to determine the average anterior-posterior (AP) length. Each knee was then scaled in three dimensions to match the average AP length. Geomagic's least-squares algorithm was used to create an average surface model. This method was validated by generating a statistical shaped model using principal component analysis (PCA) to compare to the least square's method. The averaged knee surface was used to design component system sizing schemes of 1, 3, 5, and 7 (fig 1). A further fifty arthritic knees were modeled to test the accuracy of fit for all component sizing schemes. Standard deviation maps were created using Geomagic to analyze the error of fit of the implant surface compared to the native femur surface. Results. The average shape model derived from Principal Component Analysis had a discrepancy of 0.01mm and a standard deviation of 0.05mm when compared to Geomagic least squares. The bearing surfaces showed a very close fit within both models with minimal errors at the sides of the epicondylar line (fig 2). The surface components were lined up posteriorly and distally on the 50 femurs. Statistical Analysis of the mesh deviation maps between the femoral condylar surface and the components showed a decrease in deviation with a larger number of sizes reducing from 1.5 mm for a 1-size system to 0.88 mm for a 7-size system (table 1). The femoral components of a 5 or 7-size system showed the best fit less than 1mm. The main mismatch was on the superior patella flange, with maximum projection or undercut of 2 millimeters. Discussion and Conclusion. The study showed an approach to total knee design and technique for a more accurate reproduction of a normal knee. A 5 to 7 size system was sufficient, but with two widths for each size to avoid overhang. Components based on the average anatomic shapes were an accurate fit on the bearing surfaces, but surgery to 1-millimeter accuracy was needed. The results showed that an accurate match of the femoral bearing surfaces could be achieved to better than 1 millimeter if the component geometry was based on that of the average femur. For any figures or tables, please contact the authors directly

Patellofemoral joint (PFJ) arthroplasty is traditionally performed using mechanical jigs to align the components, and it is hard to fine tune implant placement for the individual patient. These replacements have not had the same success rate as other forms of total or partial knee replacement surgery1. Our team have developed a computer assisted planning tool that allows alignment of the implant based on measurements of the patient's anatomy from MRI data with the aim of improving the success of patellofemoral joint arthroplasty. When planning a patellofemoral joint arthroplasty, one must start from the premise that the original joint is either damaged as a result of osteoarthritis, or is dysplastic in some way, deviating from a normal joint. The research aimed to plan PFJ arthroplasty using knowledge of the relationship between a normal PFJ (trochlear groove, trochlea axis and articular surfaces) and other aspects of the knee2, allowing the plan to be estimated from unaffected bone surfaces, within the constraints of the available trochlea. In order to establish a patient specific trochlea model a method was developed to automatically compute an average shape of the distal femur from normal distal femur STL files (Fig.1). For that MRI scans of 50 normal knees from osteoarthritis initiative (OAI) study were used. Mimics and 3-matic software (Materialise) packages were used for segmentation and analysis of 3D models. Spheres were fitted to the medial and lateral flexion facets for both average knee model and patient knee model. The average knee was rescaled and registered in order to match flexion facet axis (FFA) distance and FFA midpoint of the patient (Fig.2). The difference between the patient surface and the average knee surface allow to plan the patella groove alteration. The Patella cut is planned parallel to the plane fitted to the anterior surface of the patella. The patella width/thickness ratio (W/T=2) is used to predict the post reconstruction thickness3. The position of the patella component (and its orientation if a component with a median ridge is used) is also planned. The plan is next fine-tuned to achieve satisfactory PFJ kinematics4 (Fig.3). This will be complemented by intraoperative PFJ tracking which assists with soft tissue releases. PFJ kinematics is evaluated in terms of patella shift, tilt and deviation from the previously described circular path of the centre of the patella. The effect of preoperative planning on PFJ tracking and soft tissue releases is being examined. Additional study is needed to evaluate whether planning and intraoperative kinematic measurements improve the clinical outcome of PFJ arthroplasty

Introduction. Despite generally excellent patient outcomes for Total Knee Arthroplasty (TKA), there remains a contingent of patients, up to 20%, who are not satisfied with the outcome of their procedure. (Beswick, 2012) There has been a large amount of research into identifying the factors driving these poor patient outcomes, with increasing recognition of the role of non-surgical factors in predicting achieved outcomes. However, most of this research has been based on single database or registry sources and so has inherited the limitations of its source data. The aim of this work is to develop a predictive model that uses expert knowledge modelling in conjunction with data sources to build a predictive model of TKR patient outcomes. Method. The preliminary Bayesian Belief Network (BBN) developed and presented here uses data from the Osteoarthritis Initiative, a National Institute of Health funded observational study targeting improved diagnosis and monitoring of osteoarthritis. From this data set, a pared down subset of patient outcome relevant preoperative questionnaire sets has been extracted. The BBN structure provides a flexible platform that handles missing data and varying data collection preferences between surgeons, in addition to temporally updating its predictions as the patient progresses through pre and postoperative milestones in their recovery. In addition, data collected using wearable activity monitoring devices has been integrated. An expert knowledge modelling process relying on the experience of the practicing surgical authors has been used to handle missing cross-correlation observations between the two sources of data. Results. The model presented here has been internally cross validated and has some interesting facets, including the strongest single predictive question of bad outcome for the patient being the presence of lower back pain. Clinical implementation and long term predictive accuracy result collection is ongoing. Discussion. Unsatisfied patients represent a significant minority of TKR recipients, with multiple, multifaceted causal factors both in surgery and out implicated. Historically, focus has been on the role of management and improvement of the surgical factors, which is linked to the fact that surgical factors can often lead to far more disastrous consequences for the patient and the basic principle that “you only improve what you measure.” Growing collection of Patient Reported Outcome Measures by registries around the world has exposed the fact that management of patient factors has lagged behind. (Judge, 2012) Increasingly, the pivotal role of unmet expectations in determining patient satisfaction (Noble, 2006) and the “expectation gap” (Ghomrawi, 2012) between surgeons and patients has been exposed as an opportunity to improve patient outcomes. By developing a model that uses existing surgical expert knowledge to integrate research identified preoperative factors that can be accurately and practically gathered in a clinical setting, a workflow that manages patient expectations in order to optimize outcomes could reduce dissatisfaction rates in TKR recipients. Future work should focus on improving clinical integration and, in the absence of sufficiently wide, deep and complete patient response and predictor datasets, ways of harnessing existing expert knowledge into an evolving predictive tool of patient outcomes