Aims. Wear is difficult to predict in mixed lubricated articulating surfaces and the time of computation is one of the challenges due to the deterministic definition of roughness on a micro-scale. This research aims to efficiently capture the wear and the evolution of the roughness of mixed lubricated bearing surfaces, employing a statistical description of the roughness. Methods. A numerical model was developed which characterizes the wear of a loaded and lubricated pin-on-plate system, assuming a rough non-wearing pin and a rough wearing plate. The part of the load, which is borne by asperities in contact, is derived from the Greenwood-Williamson approach and the rest, which is carried by the fluid film, is based on the Patir-Cheng flow factors lubrication method. Wear is computed in the areas of direct solid contact only. For simplicity, the depth of the pin and plate are assumed infinite in order to reduce the lubrication problem to one-dimension. The roughness and asperities are described by their Cumulative Distribution Functions (CDFs). As the plate runs-in the pin, the roughness of the plate is worn by the roughness of the pin, and the process is continued until steady wear is attained. The local gap-dependent flow factors influence the load carried by the thin film of the lubricant, whereas, the local gap-dependent overlap of asperities of the pin and the plate determines the true contact load. The sum of fluid and solid contact load is balanced with the applied load, adjusting the separation between the plate and the pin. The plate asperity CDFs are updated assuming Archard's wear model for the solid contact only and the asperity wear is extrapolated to update the roughness of the plate. Results and conclusion. In this study, material properties and roughness information of silicon nitride coatings are used for the pin and the plate. The evolution of load sharing due to wear reported a decrease of asperity contact load with an increase in fluid load since the total load is balanced with the applied load. A decrease in the contact area ratio is also observed, i.e. the ratio of total contact area to the nominal area reduced due to a lesser contact load. Effect of wear on the plate roughness CDF, indicated the smoothing of asperities during prolonged wear. In addition, a friction study varying the entrainment velocities is conducted to show all the regimes of lubrication. The boundary regime had a solid friction coefficient of 0.16, followed by a decrease in friction coefficient from a value of 0.16 to 0.005 in the mixed regime and an increase to a value of 0.02 in the full film regime. The model is numerically efficient in computing wear for mixed lubricated problems, given the CDFs of two real rough surfaces

Introduction. Large-scale retrieval studies have shown backside wear in tibial inserts is dependent on the surface roughness of the tibial tray. Manufacturers acknowledge this design factor and have responded with the marketing of mirror-finished trays, which are clinically proven to have lower wear rates in comparison to historically “rough” (e.g. grit blasted) trays. While the relationship between wear and surface roughness has been explored in other polymer applications, the quantitative dependence of backside wear rate on quantitative surface finish has not yet been established for modern devices. The present study evaluates small-excursion polyethylene wear on pucks of a variety of surface roughnesses. The objective of this study is to determine where inflection points exist in the relationship between surface roughness and wear rate. Materials and Methods. An AMTI Orthopod, 6-station pin on disk tribotest was designed to mimic worst-case in vivo backside wear conditions based on published retrieval analyses. Titanium (Ti6Al4V) pucks with six different surface roughness preparations (Sa ranges from 0.06 um to 1.06 um) were characterized with white light profilometry. Never implanted polyethylene tibial inserts (never irradiated, EtO sterilized) were machined into 6 mm diameter cylindrical pins. Fretting-type motion was conducted in a 2mm square pattern at 2Hz under 100 N constant force in 25% bovine serum lubricant for 1.35 million cycles in triplicate. Mass measurements were taken every 225 thousand cycles. Results. Over the range of surface roughness studied (Sa = 0.06 – 1.06 µm), wear rate grew logistically. The wear rate for highly polished titanium (Sa = 0.06 µm) was not statistically different from less-polished titanium with Sa of 0.14 µm (p > 0.1). Titanium pucks having the highest surface roughness (Sa > 0.5µm), removed material significantly faster than those with roughness less than 0.3µm. The results of these tests suggest that Ti trays with Sa less than 0.15µm may yield equivalent clinical backside wear results, while pucks with Sa greater than 0.15µm begin to have increased wear rates that may be clinically significant. The two pucks with Sa greater than 0.5 µm yielded wear rates failing to be statistically differentiable (p = 0.059), corresponding with the flattening of the logistic curve. Discussion. These results suggest that baseplates with Sa less than 0.15 µm may ultimately yield clinically equivalent outcomes. The wear rate curve changes slope between Sa 0.14 and 0.22 µm and continues to increase across the range of surface roughnesses studied. The wear rates on rough pucks (Sa > 0.5 µm) showed high variation, reducing the ability to distinguish the two statistically (p = 0.059). Further study will better distinguish wear properties at higher surface roughnesses. Conclusion. These findings demonstrate that there may be a range of finishes between a mirror polish and grit blast that may produce clinically equivalent wear rates. This work provides justification for further study into the relationship between backside wear, baseplate tray roughness, and material choices. For any figures or tables, please contact authors directly

Introduction. In Total Knee Arthroplasty (TKA) a proper choice of the implant size is mandatory in order to guarantee the success of the prosthesis, although the tribological behavior TKA is strongly influenced by the implant design. Retrieval analysis of failed total knee prosthesis is essential to investigate the wear mechanism leading to osteolysis and loosening of the implant. Assessments from retrieval studies constitute crucial information in the effort to improve prosthesis functionality and reduce the risk of revision. The aim of the present study was to investigate the correlation among different implant sizes of retrieved TKA and patients' variables such as Body Mass Index (BMI) in terms of surface modifications and morphology change, in order to examine prosthesis properties and performances. In particular, this study can improve the understanding of the tribological behavior of total knee prosthesis and it can help the surgeon to select the best implant size of TKA considering patient's variables. Methods. Twelve retrieved total knee prostheses of the same design but with different sizes were investigated. These prostheses were all cemented, fixed and posterior stabilized. These prostheses were explanted from 12 patients after a mean of 3.2 years (from 1.1 to 7.4 years). These patients had undergone a primary TKA at our hospital between 2005 and 2010; there were 10 women and 2 men with a mean age of 68 years (ranging from 48 to 77 years) at implantation. A qualitative assessment of wear patterns and surface damages was performed on femoral components and polyethylene inserts. Roughness analyses were obtained on femoral components to assess surface modifications. Surface roughness of the metallic femoral components was performed with a contact rugosimeter. Following an internal protocol, thirty measurements were acquired from each condyle. Two roughness parameters were take into account: Ra (the Mean Roughness, i.e. the arithmetical mean value of the deviations of the roughness profile about the centre line) and Rsk. (i.e. the skewness, indicates the prevalence of peaks or valleys and quantifies the asymmetry of the profile variation from the mean line). Prostheses time in-vivo and patient details were known. Results. Results from roughness studies were compared among the specimens and then matched. In Figures 1 and 2 there are showed the Ra and Rsk values measured on the retrieved femoral prosthesis. The correlation between implant size, surface roughness and morphological alterations were found and showed that prosthesis size is a significant factor on the TKA overall performance. Moreover, differences were highlighted through the specimens characterized by different time in-vivo. Discussion. No previous study, to the best of authors' knowledge, was addressed to examine the implant size of retrieved total knee prosthesis in terms of surface and morphological modifications. The present study, however, has some limitations due to the small number of retrieved TKAs and the lack of information on the level of activity of the considered patients. This study can give a novel insight into tribological mechanisms leading to knee prosthesis failure and it can also offer indications for improving its functionality

Introduction. The corrosion of modular taper junctions in hip implants is becoming an area of increased research focus. Many design factors have been hypothesized to contribute to this kind of corrosion. The authors' previous research indicated femoral stem taper roughness may influence taper corrosion. The purpose of this study is to determine whether taper roughness significantly affects taper performance. Methods. A 2. 2. design of experiment was conducted with Ti6Al4V 12/14 taper coupons coupled with CoCrMo 12/14 taper 28mm+12 heads (DePuy Synthes, Warsaw, IN) with n=3 samples per test run for a total of 12 samples. The femoral heads and taper coupons were manufactured with “smooth” finishes ranging from R. t. 100–200 µin and “rough” finishes ranging from R. t. 900–1000 µin. Test components were assembled wet (dipped in saline solution and drained) and pressed together with a 4400 N assembly force. The assemblies were immersed in phosphate buffered saline and a potentiostat was used to maintain the potential of the specimen at −50mV vs. Ag/AgCl. Incrementally larger cyclic loads were applied vertically to the head at 3Hz until a 4000N maximum load was reached, then this cyclic load was maintained for an additional 1 million cycles. Results. The long-term average corrosion test results ranged from 0.26 to 2.98 µA among the groups. The “Rough Head – Rough Stem” (Group 1) resulted in the highest average corrosion currents of 1.53 ± 0.75 µA. The “Smooth Head – Smooth Stem” (Group 4) showed the lowest average corrosion currents of 0.20 ± 0.05 µA. ANOVA analysis revealed significant differences between the groups (p>0.05), Tukey-Kramer post-hoc analysis showed a significant difference between groups 1 and 4 only. Discussion. Femoral heads and femoral stems with a smoother taper roughness specification resulted in less measured corrosion compared to components with higher taper roughness specifications under the specified test conditions. Significance. This study demonstrates taper surface roughness is a relevant design factor which could influence taper corrosion

Introduction. Implant contamination prior to cement application has the potential to affect the cement-implant bond. the consequences of implant contamination were investigated in vitro using static shear loading with bone cement and titanium dowels of differing surface roughness both with, and without contamination by substances that are likely to be present during surgery. Namely; saline, fat, blood and oil, as a negative control. Methods. Fifty Titanium alloy (Ti-6Al-4V) dowels were prepared with two surface finishes comparable to existing stems. The roughness (Ra and Rq) of the dowel surface was measured before and after the pushout test. Four contaminants (Phosphate Buffered Saline (PBS), ovine marrow, ovine blood, olive oil) were prepared and heated to 37°C. Each contaminant was smeared on the dowel surface completely and uniformly approximately 4 minutes prior to implantation. Samples were separated into ten groups (n=5 per group) based on surface roughness and contaminant. Titanium alloy dowels was placed in the center of Polyvinyl chloride (PVC) tubes with bone cement, and equilibrated at 37°C in PBS for 7 days prior to mechanical testing. The push out test was performed at 1 mm per minute. The dowel surface and cement mantel were analyzed using a Scanning Electron Microscopy (SEM) to determine the distribution and composition of any debris and contaminates on the surface. Results. All contaminants decreased stem-bone cement interfacial shear strength. Saline produced the greatest decrease, followed by blood. The effect of fat was less pronounced and similar to that of oil likely due to the strong lipid solvent properties of the methacrylate monomer. For rough dowels, there were differences in ultimate shear strength between control and contaminated groups (p<0.001). Blood and saline groups had lower ultimate shear strength compared to fat and oil (p<0.05) (fig. 1). The ultimate shear strength for smooth samples was not significantly affected by contamination. Increasing surface roughness increased the interfacial bonding strength, even in the presence of contaminants. In control, fat and oil groups, the effect of roughness are significant (p<0.001, p<0.05 and p<0.001 respectively) (fig. 1). Scanning Electron Microscopy (SEM) showed that contaminants influence the interfacial bond by different mechanisms. Although rough surfaces were associated with higher bond strength, they also generated more debris, which could negatively affect the longevity of the implant bond (fig. 2 and fig. 3). Conclusion. The results of this study underscores the importance of keeping an implant free from contamination, and that if contamination does occur, a saline rinse may further decrease the stability of an implant. Contaminants did not significantly affect the bond strength between bone cement and smooth Ti stem, although a trend of improved properties was seen in the presence of lipid based contaminants. Therefore, the influence of contaminants is more important to the shape-closed type stem. Increasing surface roughness dramatically improved the load carrying capability of the implant-cement interface even with contaminants

Introduction. Recent clinical studies found no apparent reduction in wear using yttria-stabilized zirconia (Y-TZP) instead of cobalt chromium alloy femoral heads bearing against cross-linked UHMWPE. The purpose of this study was to compare the surface topography of retrieved Y-TZP and magnesia-stabilized zirconia (Mg-PSZ) femoral heads and evaluate the influence of time in vivo. The increase in average roughness (Ra, Sa) of Y-TZP due to phase transformation in vivo is well documented, while Mg-PSZ does not roughen or undergo phase transformation in vivo. However, the effects of phase transformation on the polarity (skewness, Ssk) of the surface of retrieved ZrO. 2. heads has not been reported. We hypothesized that phase transformation associated with the increased roughness of Y-TZP would influence skewness and thus the wear potential of the heads. Materials and Methods. Y-TZP (n = 18) and Mg-PSZ (n = 17) femoral heads were retrieved from revision THA. Heads were cleaned and scanned by optical profilometry (magnification = 10x) at three locations per specimen. After subtracting the curvature of the heads, roughness statistics (Sa, Ssk) were calculated and averaged for each specimen and then correlated to age in vivo, with p < 0.05 for significance. Results and Discussion. As expected, the roughness Sa of Y-TZP heads increased exponentially with age in vivo (p < 0.001, r. 2. = 0.766), while Mg-PSZ heads did not roughen with age (r. 2. = 0.007; Figure 1). The skewness data of Y-TZP retrievals were noisy with a weak positive correlation to age in vivo (r. 2. = 0.016), but were consistently positive (average Ssk of all Y-TZP specimens = 0.770), indicating its roughness was caused by positive features. Positive features such as raised edges have been reported to adversely affect wear rates in the lab, and combined with the increased average roughness suggests an accelerated wear potential with age in vivo. In contrast, the skewness data for the Mg-PSZ retrievals were negative (average Ssk = -1.00), indicating its roughness was caused by small negative features, with no relationship to age in vivo (r. 2. < 0.001). A surface best described by negative features would tend to entrap lubricant for better wear characteristics. Recent hip wear simulator tests reported Mg-PSZ femoral heads to exhibit significantly less wear compared with CoCr femoral heads. These in vitro data suggest that the low average roughness and negative polarity of Mg-PSZ femoral heads should provide superior wear characteristics in vivo. Conclusions. This study illustrated that Mg-PSZ zirconia ceramics remained stable and did not roughen or develop raised edges in vivo, in contrast to Y-TZP ceramic femoral heads. We believe that clinical studies with Mg-PSZ femoral heads would illustrate a significant reduction in wear of cross-linked UHMWPE liners. Future work will expand this study with additional specimens, and attempt to reduce the noise of the skewness data to more clearly discern any trends between Ssk and age in vivo

Introduction. Wear-related osteolysis continues to be a concern in the long-term outcome and survivorship of total hip arthroplasty (THA) and there continues to be an emphasis on bearing materials that exhibit improved wear profiles. Oxidized zirconium metal (Oxinium®, Smith & Nephew) was developed to reduce the amount of polyethylene wear as compared to cobalt chromium femoral heads, without the risk of brittle fracture seen with older generation ceramics. There are a limited number of retrieval studies evaluating the performance of Oxinium in THA. The aims of this study were 1) to visually assess damage on the surface of a large number of retrieved Oxinium femoral heads, 2) to measure surface roughness of scratches on the surfaces of Oxinium femoral heads, and 3) to use scanning electron microscopy (SEM) to assess the integrity of the oxidized zirconium surface in damaged areas. BIOLOX delta (CeramTec), a ceramic alternative to Oxinium, was included in this study for comparison. Methods. From 2006 to 2013, 59 retrieved Oxinium femoral heads in THAs were collected after an average time to revision surgery of 1.64 years. The mean patient age was 61.9 years, with 32 males and 27 females. Reasons for revision surgery were recurrent dislocation (24), femoral component loosening or subsidence (13), infection (9), acetabular loosening (4), periprosthethic fracture (4), acetabular malposition (2), heterotopic ossification (2), and 1 case of leg length discrepancy. The diameters of the femoral heads were 28 mm (9), 32 mm (22), 36mm (26) and 40mm (2). Three observers visually graded surface damage on all femoral heads according to the following criteria: 1) no scratches, 2) minimal damage with one to two scratches, 3) significant damage with multiple scratches. We measured the surface roughness of retrieved Oxinium and BIOLOX delta femoral heads with an interferomic profiler, and SEM to evaluate the extent of surface effacement. Results. Oxinium femoral heads explanted for recurrent dislocation showed substantially more severe damage as compared to heads retrieved during revision surgery for other reasons (p<0.001). Eighteen of 24 heads explanted for recurrent dislocation (75%) showed gross visual evidence of substantial surface damage (grade 3), compared to only 5 of the 35 explants for non-dislocation causes (Figure 1). The surface roughness of damaged Oxinium femoral heads was significantly higher (28.6× more rough) than undamaged Oxnium and 17.7× more rough than damaged Biolox delta heads (p<0.001; Figure 2). High magnification imaging showed severe damage and effacement of the oxidized zirconium layer, exposing the metal alloy underneath (Figure 3). This was confirmed by Energy Dispersive Xray Analysis (EDXA). Discussion. This study represents the first large-scale retrieval analysis of oxidized zirconium femoral heads in THA. The results show that high impact between an acetabular shell and an Oxinium femoral head during dislocation increases the surface roughness and causes substantial effacement to the oxidized zirconium layer compared with matched modern ceramic bearings. The surface damage seen to these femoral heads is of clinical concern because it has the potential to increase the wear of polyethylene liners in THA

In total hip arthroplasty (THA), aseptic loosening induced by polyethylene (PE) wear debris is the most important cause that limits the longevity of implants. Abrasive wear generated through the mechanism such that micrometer-roughened regions and small asperities on the metallic femoral heads surface locally plow through the PE cup surface. Abrasive wear results in the PE material being removed from the track traced by the asperity during the motion of the metallic femoral heads surface. For the purpose of reducing wear, alumina ceramics was introduced in Europe and Japan in 1970s. The clinical results of ceramic-on-PE bearings regarding the wear resistance have been superior to that of the metal-on-PE bearings. Compared with Co–Cr–Mo alloys, alumina ceramics is advantageous for precision machining because of its higher hardness, enable to form spherical and smooth surface. The fracture resistance of the alumina ceramics itself is related to grain size; the grain size reduction leads to the improvement of its resistance. In this study, we evaluated the roundness and the roughness of retrieved two distinct alumina ceramics having different grain size, and Co–Cr–Mo alloy heads. Fourteen retrieved alumina ceramic femoral heads; ten heads with a diameter of 28 mm made of small grain size alumina (SG-alumina; mean grain size is 3.4 μm) with clinical use for 16–28 years and four heads with a diameter of 26 mm made of extra-small grain size alumina (XSG-alumina; mean grain size is 1.3 μm) with clinical use for 14–19 years, were examined. Six retrieved Co–Cr–Mo alloy femoral heads with a diameter of from 22 to 32 mm with average clinical use for 12–28 years were examined. SG-alumina and XSG-alumina heads showed significantly lower roundness compared with Co–Cr–Mo alloy heads, due to higher precision machining [Fig. 1]. The surface roughness for the contact area of the heads increased in order of XSG-alumina, SG-alumina and Co–Cr–Mo alloy. The surface roughness of the non-contact area for all kinds of heads was lower than that for the contact area [Fig. 2]. Surface profiles of the SG-alumina and XSG-alumina showed the reentrant surface while Co–Cr–Mo alloy heads showed the protrusion surface. The roundness and roughness of the Co–Cr–Mo alloy or ceramic surface and the presence or absence of hard third-body particles correlate to the amount of abrasive PE wear. When the third-body was entrapped during the clinical use, a reentrant surface might be formed on the ceramic while protrusion surface formed on the Co–Cr–Mo alloy. The differences in clinical results may be due in part to the influence of third-body particles. The ceramic becomes more resistant than Co–Cr–Mo alloy against the scratching by the entrapped abrasive contaminants because of its harder surface. From the good clinical results of more than 20 years using SG-alumina, the greater long term clinical results using XSG-alumina will be expected

Introduction. The initial mechanical stability of cementless femoral stems in total hip arthroplasty is an important factor for stable biological fixation. Conversely, insufficient initial stability can lead to stem subsidence, and excessive subsidence can result in periprosthetic femoral fracture due to hoop stress. The surface roughness of stems with a surface coating theoretically contributes to initial mechanical stability by increasing friction against the bone, however, no reports have shown the effect of surface roughness on stability. The purpose of this study was to evaluate the effect of differences in surface roughness due to different surface treatments with the same stem design on the initial stability. Materials and Methods. Proximally titanium plasma-sprayed femoral stems (PS stem) and proximally grit-blasted stems (GB stem) were compared. The stem design was identical with an anatomic short tapered shape for proximal fixation. The optimum size of PS stem based on 3D templating was implanted in one side of 11 pairs of human cadaveric femora and the same size of GB stems was implanted in the other side. After implantation, the specimens were fixed to the jig of a universal testing machine in 25cm of entire length so that the long axis of the femur was positioned at 15-degrees adduction to the vertical. Vertical load tests were conducted under 1 mm/minute of displacement-controlled conditions. After 200 N of preload to eliminate the variance in the magnitude of press-fit by manual implantation, load was applied until periprosthetic fracture occurred. Results. The same size of PS or GB stem was successfully implanted in all 11 pairs without fracture. The distances of subsidence until fracture occurred were 2.2±1.2 mm in the PS stem and 2.5±1.1 mm in the GB stem and no significant difference was detected. The load applied for 1 mm of subsidence was 792±478 N in the PS stem and 565±431 N in the GB stem and there was a significant difference between the two groups. The load at fracture was 3037±1563 N in the PS stem and 2614±1484 N in the GB stem and there was a significant difference between the groups. Discussion. A significantly larger load was applied for 1 mm of subsidence in the PS stem compared to the GB stem. This suggests that the plasma-spray porous-coated surface had a less slippery interface than the grit-blasted surface. Both femora of a pair fractured at the same level of hoop stress that was induced by the same amount of stem subsidence but at significantly different loads. The fact that the load at fracture in the PS stem was significantly larger than that in the GB stem was due to differences in shear stress caused by different levels of friction. The scratching effect against the femoral canal due to the rougher surface of the plasma-spray porous-coating works advantageously for initial mechanical stability

Introduction. Implant-cement debonding at the knee has been reported previously [1]. The strength of the mechanical interlock of bone cement on to an implant surface can be associated with both bone cement and implant related factors. In addition to implant surface profile, sub-optimal mixing temperatures and waiting times prior to cement application may weaken the strength of the interlock. Aims. The study aimed to investigate the influence of bone cement related factors such as mixing temperature, viscosity, and the mixing and waiting times prior to application, in combination with implant surface roughness, on the tensile strength at the interface. Materials and Methods. Tensile tests were carried out on two types of hand-mixed cement, high (HV) and medium viscosity (MV), sandwiched between two cylindrical Cobalt-Chrome coupons with either smooth (60 grit) or rough (20 grit) surface finishes. 144 Specimens were prepared with a cement thickness layer of 2.5 mm in customised rigs (Figure 1). The samples were grouped and tested at two mixing temperatures (23 and 19 degrees), at different mixing times (HV-30s, MV-45s). Waiting times after mixing were varied between early (1.5 min), optimal (4.5 min) or late (8 min); for HV and 4 min, 7.5 min and 11 min for MV cements. All the samples were cured for 24 hours prior to testing. The peak force and stress was calculated for all specimens. Results and Conclusion. Surface Finish: Rough surfaced samples had significantly higher (p < 0.05) mean tensile forces and stress than smooth samples at both 19 and 23 degrees across HV and MV cement types. Cement Type: MV cements, when applied to rough samples with waiting times of 4 minutes at 23 degrees, and 11 minutes at 19 degrees, resulted in the highest peak tensile forces, followed by 7.5 minutes at 23 and 19 degrees respectively (Figure 2). Temperature at different application times for rough and smooth samples: for MV cement, rough samples prepared at 23 degrees, 4 minutes, and smooth samples at 19 degrees, 7.5 minutes were found to be significantly better (p < 0.05) than their counterparts. For HV cement, 23 degrees was found to be better (p < 0.05) for smooth samples at applications times of 4.5 and 8 minutes and 19 degrees for application times of 1.5 minutes. No significant difference was noted for rough samples for the same. Application times at different temperatures for rough and smooth samples: at both 19 and 23 degrees, there were no differences between application times within the rough sample groups for HV or MV. However, for smooth samples, HV cement, tensile forces were significantly higher (p < 0.05) at 23 degrees in the following order; 8 minutes > 4.5 minutes > 1.5. The results show that implant surface roughness and cement mixing time, temperature, viscosity and application times affect the strength of the interlock at the interface

Previous retrieval studies demonstrate increased tibial baseplate roughness leads to higher polyethylene backside wear in total knee arthroplasty (TKA). Micromotion between the polyethylene backside and tibial baseplate is affected by the locking mechanism design and can further increase backside wear. The purpose of this study was to examine modern locking mechanisms, in the setting of both roughened and polished tibial baseplates, on backside tibial polyethylene wear. Five TKA models were selected, all with different tibial baseplate and/or locking mechanism designs. Six retrieval tibial polyethylenes from each TKA model were matched based on time in vivo (TIV), age at TKA revision, BMI, gender, number of times revised, and revision reason. Two observers scored each polyethylene backside according to a visual damage score and individual damage modes. Primary outcomes were mean damage score and individual damage modes. Demographics were compared by one-way ANOVA. Damage scores and modes were analysed by the Kruskal-Wallis test and Dunn's multiple comparisons test. There were no differences among the groups based on TIV (p=0.962), age (p=0.651), BMI (p=0.951), gender, revision number, or reason for revision. There was a significant difference across groups for mean total damage score (p=0.029). The polished tibial design with a partial peripheral capture locking mechanism and anterior constraint demonstrated a significantly lower score compared to one of the roughened tibial designs with a complete peripheral-rim locking mechanism (13.0 vs. 22.1, p=0.018). Otherwise, mean total damage scores were not significant between groups. As far as modes of wear, there were identifiable differences among the groups based on abrasions (p=0.005). The polished design with a tongue-in-groove locking mechanism demonstrated a significantly higher score compared to both groups with roughened tibial baseplates (5.83 vs. 0.83, p=0.024 and 5.83 vs. 0.92, p=0.033). Only the two designs with roughened tibial baseplates demonstrated dimpling (5.67 and 8.67) which was significant when compared against all other groups (p0.99). No other significant differences were identified when examining burnishing, cold flow, scratching, or pitting. No polyethylene components exhibited embedded debris or delamination. Total damage scores were similar between all groups except when comparing one of the polished TKA design to one of the roughened designs. The other TKA model with a roughened tibial baseplate had similar damage scores to the polished designs, likely due to its updated locking mechanism. Dimpling wear patterns were specific for roughened tibial baseplates while abrasive wear patterns were identified in the design with a tongue-in-groove locking mechanism. Our study showed even in the setting of a roughened tibial baseplate, modern locking mechanisms decrease backside wear similar to that of other current generation TKA designs

Purpose. Previous retrieval studies demonstrate increased tibial baseplate roughness leads to higher polyethylene backside wear in total knee arthroplasty (TKA). Micromotion between the polyethylene backside and baseplate is affected by the locking mechanism design and can further increase backside wear. This study's purpose was to examine modern locking mechanisms influence, in the setting of both polished and non-polished tibial baseplates, on backside tibial polyethylene damage and wear. Methods. Five TKA models were selected with different tibial baseplate and/or locking mechanism designs. Six retrieval tibial polyethylenes from each TKA model were matched based on time in vivo (TIV), age at TKA revision, BMI, gender, number of times revised, and revision reason. Two observers visually assessed each polyethylene. Primary outcomes were visual damage scores, individual visual damage modes, and linear wear rates determined on micro-computed tomography (micro-CT) scan in mm/year. Demographics were compared by one-way ANOVA. Damage scores, damage modes, and linear wear were analyzed by the Kruskal-Wallis test and Dunn's multiple comparisons test. Results. There were no differences among the groups based on TIV (p=0.962), age (p=0.609), BMI (p=0.951), gender, revision number, or reason for revision. There was a significant difference across groups for visual total damage score (p=0.031). The polished tibial design with a partial peripheral capture locking mechanism and anterior constraint demonstrated a significantly lower score compared to one of the non-polished tibial designs with a complete peripheral-rim locking mechanism (13.0 vs. 22.0, p=0.019). Otherwise, mean total damage scores were not significant between groups. There were identifiable differences among the groups based on abrasions (p=0.006). The polished design with a tongue-in-groove locking mechanism demonstrated a significantly higher score compared to one of the designs with a non-polished baseplate (5.83 vs. 0.83, p=0.016). Only the two designs with non-polished baseplates demonstrated dimpling (5.67 and 8.67), which was significant when compared against all other groups (p<0.0001), but not against each other (p>0.99). No other significant differences were identified when examining burnishing, cold flow, scratching, or pitting. No polyethylene components exhibited embedded debris or delamination. There was a significant difference among groups for linear wear on micro-CT scanning (p=0.003). Two of the polished baseplate designs, one with the partial peripheral rim capture and one with the tongue-in-groove locking mechanism, demonstrated significantly lower wear rates than the non-polished design with a complete peripheral-rim locking mechanism (p=0.008 and p=0.032, respectively). There were no other differences in wear rates between groups. Conclusions. Total damage scores and wear rates were similar between all groups except when comparing two of the polished TKA designs to one of the non-polished baseplate designs. The other TKA model with a non-polished tibial baseplate had similar damage scores and wear rates to the polished designs, likely due to its updated locking mechanism. Dimpling was specific for non-polished tibial baseplates while abrasions were identified in the design with a tongue-in-groove locking mechanism. Our study showed even in the setting of a non-polished tibial baseplate, modern locking mechanisms can decrease backside damage and wear similar to that of other current generation TKA designs. For any figures or tables, please contact authors directly (see Info & Metrics tab above).

The design of the femoral prosthesis in cementless total hip arthroplasty is known to affect the initial strains in the cortex during implantation and in the early postoperative time period. High strains have a direct influence on periprosthetic fracture. This study compares the existing ABGII stem, which is proximally coated with a grit blasted titanium surface with hydroxyapatite coating with a prototype that has a rougher titanium plasma spray proximal coating. The Australian National Joint registry results 2011 reported the ABG2 femoral component cumulative percent revision (CPR) of 6.5 (93.5% survival), which compares favourably with equivalent stems with 10 year CPR data such as the Taperloc 6.6 and Corail 7.3. Six pairs of fresh-frozen cadaveric femurs were mounted in blocks according to ISO guidelines in single leg stance setup. Five strain gauges were attached around the neck of the femur and then prepared according to routine operative techniques to accept the femoral prosthesis. Cortical strains were measured during insertion of the prosthesis with an instrumented mallet attached to an accelerometer. Subsequently, force-displacement readings were taken during cyclical loading on a servo-hydraulic machine and finally the stems were tested to failure. Our results showed significantly less strain during cyclical loading of the stem with increased surface roughness (p < 0.05). They also showed no significant differences loads/strains during impaction (p = 0.159), no significant difference in micromotion (p = 0.148) and no significant difference in load-to-failure (p = 0.37)

Introduction. The release of metallic debris can promote many adverse tissue reactions, as metallosis, necrosis, pseudotumors and osteolysis . 1–3. This debris is mainly generated by the fretting-corrosion mechanism due to the geometric difference in the head-stem interface . 4. Retrieval and in silico analysis showed the roughness of the stem-head interface appears to play an important role in the volume of material lost and THA failure . 5–7. The technical standard ISO 7206-2 recommends the measurement of average roughness (Ra) and max height of the profile (Rz) to control the quality of the surface finish of articulating surfaces on THA implants. However, despite the importance of the trunnion roughness, there is no specific requirement for this variable on the referred technical standard. The present study carried out a surface finish analysis of the trunnion of hip stems from five distinct manufacturers. Methods. Four stems (n = 4) from five (5) distinct manufacturers (A, B, C, D, and E) were used to evaluate the roughness of the trunnion. All the stems are similar to the classical Exeter stem design, with a 12/14 taper and a polished body surface. The roughness of trunnions was evaluated according to ISO 4287 and ISO 13565-2. The total assessment length was 4.8 mm with 0.8 mm cut-off. The first and last 8.33% of assessment length were not considered. The measurements of all samples were made in a rugosimeter with 2 µm feeler ITP (Völklingen, Germany), the velocity of 0.5 mm.s. -1. , and a force of 1.5 mN. The calibration was made at 20 ºC and relative humidity at 50%. The Kruskal Wallis with post hoc Nemenyi test was used to evaluate the difference of Ra among the manufacturers. The confidence level was set at 5%. Results and Discussion. The analysis of surface finish revealed different roughness among the manufactures (p < 0.005), with Ra between 0.061 µm to 3.184 µm and Rz varying of 0.41 µm to 12.69 µm. The manufacturers A and E had a Ra (2.587±0.050 µm and 3.146±0.031µm) of the trunnion similar to founded by Panagiotidou et al (2013). Within such range, the trunnion has shown a high presence of pit . 8. The manufacturer C, on the other hand, had the best surface finish of the trunnion (Ra = 0.069±0.010 µm and Rz = 0.505 ± 0.076 µm). This more smooth surface might increase the taper strength, reduce the shear stress and the susceptibility to the fretting-corrosion damage . 4,8. . Conclusion. The results were worrying because there is great variability of roughness among the manufacturers with the occurrence of trunnions with roughness too high. Nevertheless, the ISO technical standard does not recommend any procedure or minimum parameters acceptable for the surface finish of the trunnion. The revision of ISO 7206-2 would guarantee better control of trunnion roughness to reduce the amount of metallic debris and increase the safety of THA implants. Additional research is needed to determine a target value for this variable. For any figures or tables, please contact authors directly

Introduction. The accumulation of proteins and bacteria on implant surfaces is a critical concern in the biomedical field, especially with respect to the potential of biofilm formation on implant surfaces. Material surface wettability is often used as a predictor of potential colonization of specific bacterial strains. Surface roughness has also been shown to have a strong relationship with biofilm formation, as rougher surfaces tend to have a stronger affinity to harbor bacterial colonies. The modification of implant surfaces to impart a biofilm resistant layer can come at the expense of increasing surface roughness however, and it is therefore important to determine how the variables of wettability and roughness are affected by any new surface coating technologies. In the current work, a novel CoBlast (C) process that impregnates alumina (A) at 50 μm grit (5) or 90 μm grit (9) sizes, with the possible addition of polytetrafluoroethylene (P) onto titanium surfaces, combined with a plasma coating process called BioDep, that coats the surface with chitosan (X) with the possible addition of vancomycin (V), were evaluated for wettability and surface roughness to determine their potential as biofilm resistant treatments on implants. Materials and Methods. N=65 titanium alloy samples (n=5 for 13 sample modification types as described above and in the figure legends below) were analyzed for surface roughness and wettability. Following cleaning in ethanol, roughness testing (Ra, Rq, Rt and Rz, Wyko NT-2000 optical profilometer @ 28.7× magnification, FOV of 164×215 μm) at 5 different surface locations per specimen, and contact angle analysis was performed (2 μL water drops, KRUSS EasyDrop). Statistical differences between groups was determined using ANOVA. Results and Discussion. Figure 1a summarizes the roughness results, with significant roughening being observed with between surface blanks and all surface modification techniques, especially the CoBlasted 90 μm grit treatments. As expected, wettability (shown in Figure 1b) was significantly affected by PTFE modifications and also by the introduction chitosan and vancomycin. Conclusions. As can be seen from these results, changing the coating of a material can change the surface topography and the wettability of the surface, which can be beneficial for different applications. The results from this work show that the CoBlast and BioDep processes significantly affect both wettability and roughness, and that the benefits and potential drawbacks of each must be considered when assessing their potential for biofilm resistance. PTFE-coated samples would be best used when wanting to prevent a hydrophobic substance from binding to the material, while the alumina-coated or blank samples would be best used to prevent a hydrophilic substance from binding. In the future, nonpolar liquid wettability will be assessed to better mimic in-vivo conditions and to determine surface energy to be able to make better conclusions about the relationship between surface roughness and wettability. For any figures or tables, please contact the authors directly

Introduction. Metal-on-polyethylene (MoP) is the most commonly used bearing couple in total hip replacements (THRs). Retrieval studies (Cooper et al, 2012, JBJS, Lindgren et al, 2011, JBJS) report adverse reactions to metal debris (ARMD) due to debris produced from the taper-trunnion junction of the modular MoP THRs. A recent retrospective observational study (Matharu et al, 2016, BMC Musc Dis) showed that the risk of ARMD revision surgery is increasing in MoP THRs. To the authors' best knowledge, no hip simulator tests have investigated material loss from the taper-trunnion junction of contemporary MoP THRs. Methods. A 6-station anatomical hip joint simulator was used to investigate material loss at the articulating and taper-trunnion surfaces of 32mm diameter metal-on-cross-linked polyethylene (MoXLPE) joints for 5 million cycles (Mc) with a sixth joint serving as a dynamically loaded soak control. Commercially available cobalt-chromium-molybdenum (CoCrMo) femoral heads articulating against XLPE acetabular liners (7.5Mrad) were used with a diluted new-born-calf-serum lubricant. Each CoCrMo femoral head was mounted on a 12/14 titanium alloy trunnion. The test was stopped every 0.5Mc, components were cleaned and gravimetric measurements performed following ISO 14242-2 and the lubricant was changed. Weight loss (mg) obtained from gravimetric measurements was converted into volume loss (mm. 3. ) and wear rates were calculated from the slopes of the linear regression lines in the volumetric loss versus number of cycles plot for heads, liners and trunnions. Additionally, volumetric measurements of the head tapers were obtained using a coordinate measuring machine (CMM) post-test. The surface roughness (Sa) of all heads and liners was measured pre and post-test. At the end of the test, the femoral heads were cut and the roughness of the worn and unworn area was measured. Statistical analysis was performed using a paired-t-test (for roughness measurements) and an independent sample t-test (for wear rates). Results and Discussion. The mean volumetric wear rates for CoCrMo heads, XLPE liners and titanium trunnions were 0.019, 2.74 and 0.013 mm. 3. /Mc respectively. There was a statistically significant decrease (p<0.001) in the Sa of the liners post-test. This is in contrast to the femoral heads roughness in which no change was observed (p = 0.338). This head roughness result matches with a previous MoP in vitro test (Saikko, 2005, IMechE-H). The Sa of the head tapers on the worn area showed a statistically significant increase (p<0.001) compared with unworn, with an associated removal of the original machining marks. The mean volumetric wear rate of the head tapers obtained using the CMM (0.028 ± 0.016 mm. 3. /Mc) was not statistically different (p=0.435) to the mean volumetric wear rate obtained gravimetrically (0.019 ± 0.020 mm. 3. /Mc) for the femoral heads. Therefore, wear of the heads arose mainly from the internal taper. The mean wear rates of the CoCrMo taper and titanium trunnion are in agreement with a MoP explant study (Kocagoz et al, 2016, CORR). Conclusion. This is the first long-term hip simulator study to report wear generated from the taper-trunnion junction of MoP hips

Introduction. Total ankle replacement (TAR) has been used as a surgical intervention for arthritis since the 1970s. However, unlike clinically successful hip and knee replacements, TARs are renowned for extensive contraindications to surgery and high failure rates with an average of 83% survival at 5 years. The majority cite aseptic loosening as the reason for failure. The aim of this study wais to analyse retrieved TARs visually and through interferometry to identify potential the failure mechanisms associated with these devices. Methods. Retrieved total ankle replacements (n=11) from consecutive revision surgeries carried out at Chapel Allerton Hospital, Leeds between August 2012 and January 2014, were collected for study at the University of Leeds, under an NRES approved procedure (09/H1307/60). The bearing surfaces of the samples were visually inspected for evidence of damage and wear. The bearing surfaces between the tibial component and the flat surface of the polyethylene insert were then examined using a scanning white light interferometer (NP Flex, Bruker, USA). It was not possible to characterise the talar bearing surface or the inferior polyethylene surface at this stage through interferometry due to the curvature of the surface. The components were aligned and five sections on each of the surfaces measured. These sections represented; anterior-medial, anterior-lateral, posterior-medial, posterior lateral and central regions of the bearing surfaces. 3D roughness values were recorded, and the mean 3D surface roughness compared between implants. Measurements were taken on the medial and lateral aspects of the bearing surfaces to investigate whether damage was location specific. A coefficient of determination was calculated to assess the relationship between implantation time and surface roughness. Results. Eleven total ankle replacements were received, with six right and five left TARs (Figure 1). The mean implantation time was 7.0 ± 1.4 years. Ten bearings were revised due to osteolytic cysts. Visual inspection showed evidence of multidirectional motion on both the polyethylene and metallic tibial bearing surfaces for all retrievals. Evidence of scratching, pitting, burnishing, adhesive wear was seen on all inserts. There was no evidence of gross polyethylene wear. Visual inspection indicated some metallic particles embedded in the polyethylene. The 3D surface roughness (Sa) ranged from 16.7 to 53.1nm across the tibial components. The roughness of the polyethylene surface ranged from 32.1nm to 2312.6nm (Figure 2). Such high roughness and the visible 20μm pits indicate the presence of third body particle between the articulating surfaces. There was a weak positive correlation between implantation time and surface roughness of the insert (R2=0.33) but no correlation for the tibial bearing. Discussion. Visual inspection of the retrieved samples showed extensive surface damage to both the tibial bearing and the polyethylene insert. Presence of metallic debris embedded in the insert indicated potential for third body wear. A weak correlation indicated that bearings implanted for the longest duration had the greatest surface roughness. Whilst no clear failure mechanisms have been determined through this initial analysis, this is the first study to assess surface characteristics of failed total ankle replacements

Introduction. The ability to create patient-specific implants (PSI) at the point-of-care has become a desire for clinicians wanting to provide affordable and customized treatment. While some hospitals have already adopted extrusion-based 3D printing (fused filament fabrication; FFF) for creating non-implantable instruments, recent innovations have allowed for the printing of high-temperature implantable polymers including polyetheretherketone (PEEK). With interest in FFF PEEK implants growing, it is important to identify methods for printing favorable implant characteristics such as porosity for osseointegration. In this study, we assess the effect of porous geometry on the cell response and mechanical properties for FFF-printed porous PEEK. We also demonstrate the ability to design and print customized porous implants, specifically for a sheep tibial segmental defect model, based on CT images and using the geometry of triply periodic minimal surfaces (TPMS). Methods. Three porous constructs – a rectilinear pattern and gyroid/diamond TPMSs – were designed to mimic trabecular bone morphology and manufactured via PEEK FFF. TPMSs were designed by altering their respective equation approximations to achieve desired porous characteristics, and the meshes were solidified and shaped using a CAD workflow. Printed samples were mCT scanned to determine the resulting pore size and porosity, then seeded with pre-osteoblast cells for 7 and 14 days. Cell proliferation and alkaline phosphatase activity (ALP) were evaluated, and the samples were imaged via SEM. The structures were tested in compression, and stiffness and yield strength values were determined from resulting stress-strain plots. Roughness was determined using optical profilometry. Finally, our process of porous structure design/creation was modified to establish a proof-of-concept workflow for creating PSIs using geometry established from segmented sheep tibia CT images. Results. ALP activity measurements of the porous PEEK samples at 7 and 14 days were significantly greater than for solid controls (p < 0.001 for all three designs, 14 days). No difference between the porous geometries was found. SEM imaging revealed cells with flat, elongated morphology attached to the surface of the PEEK and into the pore openings, with filopodia and lamellipodia extensions apparent. mCT imaging showed average pore size to be 545 ± 43 µm (porosity 70%), 708 ± 64 µm (porosity 68%), and 596 ± 94 µm (porosity 69%) for the rectilinear, gyroid, and diamond structures, respectively. The average error between the theoretical and actual values was −16.3 µm (pore size) and −3.3 % (porosity). Compression testing revealed elastic moduli ranging from 210 to 268 MPa for the porous samples. Yield strengths were 6.6 ± 1.2 MPa for lattice, 14.8 ± 0.7 MPa for gyroid, and 17.1 ± 0.6 for diamond. Average roughness ranged from 0.8 to 3 µm. Finally, we demonstrated the ability to design and print a fully porous implant with the geometry of a sheep tibia segment. Assessments of implant geometrical accuracy and mechanical performance are ongoing. Discussion. We created porous PEEK with TPMS geometries via FFF and demonstrated a positive cellular response and mechanical characteristics similar to trabecular bone. Our work offers an innovative approach for advancing point-of-care 3D printing and PSI creation

Fixation of cemented femoral stems is reproducible and provides excellent early recovery of hip function in patients 60–80 years old. The durability of fixation has been evaluated up to 20 years with 90% survivorship. The mode of failure of fixation of cemented total hip arthroplasty is multi-factorial; however, good cementing techniques and reduction of polyethylene wear have been shown to reduce its incidence. The importance of surface roughness for durability of fixation is controversial. This presentation will describe my personal experience with the cemented femoral stem over 30 years with 3 designs and surface roughness (RA) ranging from 30 to 150 microinches. Results. Since 1978, three series of cemented THA have been prospectively followed using periodic clinical and radiographic evaluations. All procedures were performed by the author using the posterior approach. Excellent results and Kaplan-Meier survivorship ranges from 90% to 99.5% in the best case scenario were noted at 10 to 20 year follow-up. Conclusion. With a properly-designed femoral stem, good cement technique, proper cement mantle, and surface roughness of 30 to 40 microinches, the cemented femoral stem provides a durable hip replacement in patients 60 to 80 years old with up to 95% survivorship at 10 to 20 year follow-up

Introduction. Cementless devices can be designed with varying surface treatments with the hope of achieving osseointegration. The surface finish dictates the interaction, adhesion and growth of bone therefore it is an important parameter that be measured and compared. The surface topography of a material can be viewed both microscopically and macroscopically. Surface microtopography focuses on the peaks and valleys where deviations in the characteristics of the size and spacing of these features determine the variability between surface topography. The most common parameter used worldwide to describe surface roughness is the arithmetic average height (Ra). The definition of Ra is the absolute deviation of the surface irregularities from a mean line across the sampling length given by the equation shown in figure 1. Many techniques can be used to relate to surface characteristics of materials, with the common two options revolving around contact and non-contact methods. These techniques are expensive and are limited in detecting the interaction of implantable devices at a macroscopic level. This study sought to develop a method to determine the surface roughness and characterise implants based on cross sectional images and scanning electron microscopy. Method. The profile of 6 trunnions from a total hip replacement was obtained in x and y coordinates along a set length using a profilometer. A custom program to calculate the Ra of the material was created using a mathematical program (MATLAB). Each material profile was inputted into a mathematical program to provide the surface roughness of the material. The surface parameters were initially obtained from a surface analyser to determine the accuracy of the program. Results. The mean difference in the Ra measurements was 0.2μm ± 0.01μm. Conclusion. This study has shown that the roughness parameters of a material obtained using the custom program was within acceptable variation to the results from the profilometer. This is indicative of the program functioning as required. The surface characteristics of implantable devices are important factors which can influence cellular behaviour hence adhesion which directly correlate to the interaction with bone. Therefore having a method to determine the surface roughness is a vital tool, which can provide insight on the inherent topography