Biotribology in Arthroplasty: Worn Surfaces Investigation on Ceramic Hip Femoral Heads Considering Wettability

: Ceramic-on-ceramic bearings for total hip replacement are considered the best choice to avoid problems such as osteolysis and wear, mainly related to soft bearings. The aim of this work was to investigate in a comparative way different kinds of ceramic femoral heads for total hip replacements from a biotribological point of view, discussing the results obtained in terms of topographies, presence of metal transfer (MT) phenomena, and wettability on their worn surfaces in a tribological framework. Different ceramic femoral heads derived from in vitro wear tests, retrieved form patients, and brand new total hip replacements were investigated. The patients group had an average age of 60 years (ranging from 27 to 83). In most cases, the cause of failure was aseptic loosening of the acetabular component. Roughness analyses were performed to measure the tribological surface evolution of the material; an SEM and EDS investigation on the explanted heads proves and quantified MT, while the wettability was measured through a novel optical laboratory set-up with the aim to furnish useful data in the framework of synovial lubrication phenomena acting in the tribosystem. For the average roughness measurements on explanted specimens were considered three parameters (Ra = the average area between the roughness profile and its mean line; Rt = the vertical distance from the deepest valley to the highest peak of the roughness profile; and Rsk = it is the skewness and it is a measure of the asymmetry of the amplitude distribution function. In other words, the skewness indicates whether a surface is dominated by peaks or by valleys) and their values were: Ra 0.22 ± 0.12 µ m, Rt 34.5 ± 13.5 µ m and Rsk − 0.01 ± 11.3; on the new specimens we measured Ra 0.01 ± 0.001 µ m, Rt 0.12 ± 0.09 µ m, and Rsk = 5.67 ± 8.7; for the in vitro specimens they were Ra 0.05 ± 0.12 µ m, Rt 0.71 ± 1.4 µ m and Rsk 7.73 ± 20.6. The wettability angle measurements showed hydrophilic surfaces for all femoral heads considered in this study with small differences between the three investigated categories, allowing to discuss their effects on the biobearings’ lubrication phenomena.


Introduction
Total hip replacement (THR) is a successful procedure with relatively low complications [1,2]. With the improvement of fixation and implant designs, and the introduction of minimally invasive techniques, the goal of THR is to minimize wear and osteolysis significantly reducing loosening of the components. Metal-on-metal (MOM) and ceramic-on-ceramic (COC) bearings are the most suitable solutions, especially for younger patients [3].
Metal-on-metal hip implants, introduced by Wiles and McKee-Farrar [4], were abandoned because of high frictional torques, inadequate surface finish, and high clinical failure rates, which resulted in In this study the authors examined 53 femoral heads. Twenty-four of them derived from explanted prosthesis which were partially investigated in a previous work of the authors [12]. The patients had undergone a primary THA at Rizzoli hospital; with an average age of 60 years (ranging from 27 to 83) at revision. In most cases, the cause of failure was aseptic loosening of the acetabular component. Prosthesis retrieved for catastrophic failures were excluded from the study, in order to avoid possible influences of other parameters on wear analysis. Major details are given in Table S1-Supporting information.
Nine femoral heads were brand new Biolox ® Delta femoral heads. The remaining 20 femoral heads were Biolox ® Forte and were analyzed after in vitro tests in a hip joint simulator realizing a hard-on-soft contact with a UHMWPE acetabular liner. In particular, these bearings were tested under bovine calf serum as lubricant, applying a simplified load as recommended by the international guidelines (ISO 14241-3000 N), and for two million cycles. All the femoral heads analyzed in this study had a diameter of 32 mm.

Surface Characterization
Measurements of surface roughness of the femoral heads were acquired using a contact rugosimeter (Hommel Tester T8000 machine, Schwenningen, Germany). Three roughness parameters (Ra, Rt, Rsk) were measured following an established protocol [18]. In details, the measurements were taken along 5 mm lines with 0.08 mm cut-off, with a length for each measurement of 0.45 mm. Each head was analyzed at four different inclinations of the stem with respect to the vertical axis, corresponding to 0 • , 15 • , 30 • and 45 • . In each angle of inclination the roughness was derived in three different points, separated by 120 • of rotation along the stem axis.
A scanning electron microscope (SEM, ZEISS EVO 50EP, Cambridge, UK) operating at 20 kV was used to characterize the surface of the explanted femoral heads. All the specimens were observed in an environmental pressure mode of 70-90 Pa in chamber, so there was no need for surface coatings. In addition, energy dispersive spectroscopy (EDS) X-ray analysis (Inca Energy-200, Oxford Instruments, High Wycombe, UK) was used to analyze their chemical composition to characterize each element and discriminate between metal scars on the surfaces.

Wettability Set-Up
Wettability is usually defined as the ability of a liquid to spread over a surface [25]. The lubricant deposited on the solid surface under gravity tends to spread until the internal cohesion forces of the liquid are balanced with the forces rising from the surface tension [26]. To evaluate the wettability of the investigated ceramic femoral heads, the apparent contact angle was evaluated applying a static drop technique. The contact angle, θ, as depicted in Figure 1, is the angle at which the liquid-vapor interface meets the solid-liquid interface. The droplet on the solid surface under gravity spreads until the liquid internal cohesive forces, the surface tension forces and gravity are in balance. Once the equilibrium is reached (three phases minimum of the energy state) a contact angle θ between the solid surface and the liquid droplet can be measured. Static contact angle θ is greater than 90 • when the surface energy of the solid-air interface assumes greater values than the liquid-air interface. With reference to water droplets (other liquids), we assumed a super-hydrophobic (super-amphiphobic) behavior for θ greater than 150 • , hydrophobic (amphiphobic) behavior for θ between 90 • and 150 • , hydrophilic (amphiphilic) behavior for θ between 10 • and 90 • , and super-hydrophilic (super-amphiphilic) for θ between 0 • and 10 • [27]. The adopted experimental apparatus consisted of a horizontal stage used to place the specimen, which allowed to be adjusted in the z-directions, a micrometer syringe to form a liquid drop, a halogen and intensity adjustable light source to illuminate the simple, and a Nikon camera for the high-resolution images acquisition ( Figure 2).
Before measuring, the samples were cleaned in an ultrasonic bath for 15 min to remove any residues, dried with nitrogen gas, and cleaned with acetone. A liquid droplet of distilled water with a volume of 10 µL was used. The liquid droplet fell on the same dome surface of the ceramic femoral heads. After camera acquisitions, the contact angle between the drop and the surface of the samples was measured using a free and open-source graphics editor (GIMP 2.8, GNU Image Manipulation Program). Tests were conducted at room temperature (25 • C); each experiment was repeated four times and the experimental results were taken as the mean value to assure reproducibility. This technique represents a novel measurement approach, used nowadays in several fields of accurate dimensional measurements, but not yet consolidated in a standardized measurement protocol, and with a complex uncertainty determination; the possibility to obtain very high magnifications of the droplet pictures at high resolution, due to the high resolution camera adopted, lent satisfactory confidence in the obtained results. The adopted experimental apparatus consisted of a horizontal stage used to place the specimen, which allowed to be adjusted in the z-directions, a micrometer syringe to form a liquid drop, a halogen and intensity adjustable light source to illuminate the simple, and a Nikon camera for the high-resolution images acquisition ( Figure 2). Before measuring, the samples were cleaned in an ultrasonic bath for 15 min to remove any residues, dried with nitrogen gas, and cleaned with acetone. A liquid droplet of distilled water with The adopted experimental apparatus consisted of a horizontal stage used to place the specimen, which allowed to be adjusted in the z-directions, a micrometer syringe to form a liquid drop, a halogen and intensity adjustable light source to illuminate the simple, and a Nikon camera for the high-resolution images acquisition ( Figure 2). Before measuring, the samples were cleaned in an ultrasonic bath for 15 min to remove any residues, dried with nitrogen gas, and cleaned with acetone. A liquid droplet of distilled water with Statistical analyses were performed on measured data by using SPSS 14.0 (SPSS Inc, Chicago, IL, USA) software. Mann-Whitney U tests and median tests with independent samples were used to check differences in distribution and in median. Statistical significance was set at p < 0.05. The correlation between roughness measurements vs. follow-up and wettability of the heads were calculated using Pearson's r.

Microscopic Results
Roughness parameters values are summarized in Table S2, Supporting Information. By analyzing the obtained data collected in Table S2, Supporting Information, it is possible to underline that the mean values of Ra, Rt and Rsk are different between the considered specimens. In fact, for the in vivo specimens the values were Ra 0.22 ± 0.12 µm, Rt 34.5 ± 13.5 µm and Rsk −0.01 ± 11.3; on the new specimens, we measured Ra 0.01 ± 0.001 µm, Rt 0.12 ± 0.09 µm, and Rsk = 5.67 ± 8.7; for the in vitro specimens, they were Ra 0.05 ± 0.12 µm, Rt 0.71 ± 1.4 µm and Rsk 7.73 ± 20.6.
A detailed comparison between considered roughness parameters and measurement angles angle is shown in Figure 3a-c, with to explanted and in vitro femoral heads.
SEM observation of some femoral heads revealed an annular wear zone where the majority of the scratches, holes and micropits were situated. The micropits and holes were generally regular in shape but varied in size (2-25 µm).
As a representative example, Figure 4 shows the picture of the microanalyses performed on one Biolox ® Delta femoral head, in which MT phenomena were observed. MT markings may consist of titanium (Ti) or cobalt chromium (CoCr) alloy and it is represented in the right part of the above picture. We observed in this part some elements, such us Ti, CoCr, Fe, Mn, which are the constituents of the metallic alloys used in the metal back.
droplet pictures at high resolution, due to the high resolution camera adopted, lent satisfactory confidence in the obtained results.
Statistical analyses were performed on measured data by using SPSS 14.0 (SPSS Inc, Chicago, IL, USA) software. Mann-Whitney U tests and median tests with independent samples were used to check differences in distribution and in median. Statistical significance was set at p < 0.05. The correlation between roughness measurements vs. follow-up and wettability of the heads were calculated using Pearson's r.

Microscopic Results
Roughness parameters values are summarized in Table S2, Supporting Information. By analyzing the obtained data collected in Table S2, Supporting Information, it is possible to underline that the mean values of Ra, Rt and Rsk are different between the considered specimens. In fact, for the in vivo specimens the values were Ra 0.22 ± 0.12 μm, Rt 34.5 ± 13.5 μm and Rsk −0.01 ± 11.3; on the new specimens, we measured Ra 0.01 ± 0.001 μm, Rt 0.12 ± 0.09 μm, and Rsk = 5.67 ± 8.7; for the in vitro specimens, they were Ra 0.05 ± 0.12 μm, Rt 0.71 ± 1.4 μm and Rsk 7.73 ± 20.6.
A detailed comparison between considered roughness parameters and measurement angles angle is shown in Figure 3a-c, with to explanted and in vitro femoral heads.
SEM observation of some femoral heads revealed an annular wear zone where the majority of the scratches, holes and micropits were situated. The micropits and holes were generally regular in shape but varied in size (2-25 μm).
As a representative example, Figure 4 shows the picture of the microanalyses performed on one Biolox ® Delta femoral head, in which MT phenomena were observed. MT markings may consist of titanium (Ti) or cobalt chromium (CoCr) alloy and it is represented in the right part of the above picture. We observed in this part some elements, such us Ti, CoCr, Fe, Mn, which are the constituents of the metallic alloys used in the metal back.

Wettability Results
The wettability apparent angle measurements, for each femoral head, are shown in Tables 1 and 2. In particular, Table 1 gives the wettability angles of the femoral heads explanted. Table 2 gives the wettability angle for each new and in vitro femoral head.

Wettability Results
The wettability apparent angle measurements, for each femoral head, are shown in Tables 1 and 2. In particular, Table 1 gives the wettability angles of the femoral heads explanted.   Table 2 gives the wettability angle for each new and in vitro femoral head. Figure 5 shows the comparison of the mean value of the measured wettability angles for each examined category; we highlight the higher wettability angle of brand new Biolox ® Delta head surfaces compare to the in vitro tested and explanted femoral heads which allow us to assert a general wear contribution on the wettability of the hip surfaces.
Regarding the observed MT, it was found that there was a greater mean MT area percentage in the case of Biolox ® Forte (5.7%) compared to Biolox Delta retrievals (4.9%).

#52
Biolox ® Forte 46 #53 Biolox ® Forte 48 Figure 5 shows the comparison of the mean value of the measured wettability angles for each examined category; we highlight the higher wettability angle of brand new Biolox ® Delta head surfaces compare to the in vitro tested and explanted femoral heads which allow us to assert a general wear contribution on the wettability of the hip surfaces.

Discussion
From a biological point of view, a high surface wettability is desirable because it promotes the material-tissue interactions [28]. From the corrosion point of view, more hydrophobic surfaces are desirable because the interaction with the external environment is minimized. Wettability can be seen as a characteristic to be correlated to the material tendency to corrode inside the human body [16,29]. In this study we investigated changes in the roughness surface in different in vitro and in vivo ceramic femoral heads, considering also MT and surface wettability. We found an average roughness measured on explanted specimens greater than the new ones and the in vitro specimens, as shown in the results section. Measurements of the observed wettability angle showed hydrophilic surface with quite high wettability for all femoral heads considered in this study, with a variation of the contact angle θ between values of 37.20 • to 55.63 • in the case of retrieved femoral heads, 33 • and 60 • in the case of new Biolox ® Delta femoral heads and 34 • to 61 • in the cases of in vitro simulated Biolox ® Forte Femoral Heads.
No statistical significance (p > 0.05) was observed between the wettability vs. the roughness measurements of all the whole femoral heads studied in this work. No statistical significance (p > 0.05) was observed between the follow-up (fu) vs. the roughness measurements of all the whole femoral heads. A statistical significance (Pearson correlation, p = 0.472) was observed between the fu vs. the Rt roughness measurements for the explanted femoral heads. Statistical significance (Pearson correlation, p = 0.706) was also found on the whole femoral heads with regard to the wettability vs. the Rt roughness measurements. A very strong correlation was found between follow-up and roughness measurements for all the explanted femoral heads (p < 0.0001).
No differences were found by comparing the percentage values of the MT relative to the Biolox ® Forte (5.38%) and Delta (5.29%) femoral heads.
It is known that roughness should act on the wettability apparent contact angle, which is not the same as the local one (the contact angle on an ideal mirror-polished surface) since different local equilibrium states on surface topographic asperities are allowed [26].
A particular importance was attributed to these results since friction and wear processes are known to be affected by the surface wettability of the prosthetic materials [16], even if definite correlations between wettability and tribological behavior are still to be established [17]. Biolox ® Delta femoral heads can be coupled with a ceramic or a polyethylene counterbody. As is known from literature, better lubrication-reduced friction and wear-is realized when the tribosystem is composed by hydrophilic materials [30,31]. The prosthetic joint lubrication can be referred to the three classical lubrication modes (boundary, mixed, full film/elastohydrodynamic) depending on the loading and the motion conditions [32]. When the specific film thickness λ, depending on the synovial film thickness and on the contact surface roughness, is large (λ > 3), the lubricating gap could ensure a complete separation between the surface asperities allowing the hydrodynamic lubrication mode [14]. When λ < 1, boundary lubrication is expected with contact between the surface asperities [15]. The transitional mixed lubrication mode occurs for 1 < λ < 3. The hydrodynamic lubrication modelling is commonly approached by the Reynolds equation under its classical hypothesis. Of course the Reynolds model should be improved considering the non-Newtonian behavior of the synovial lubricant [33,34], allowing fluidodynamic lubrication models closer to the tribosystem's real behavior.
In the above scenario, wettability plays a key role in all cases: in hydrodynamic lubrication, high wettability (and low contact angles) are often desirable since the affinity between the lubricated surfaces and the synovial fluid could improve film stability making it easy for the lubricating fluid to penetrate small gaps between surfaces. When the load and motion conditions of the prosthetic joint induce a fluid film, rupture and a boundary/mixed lubrication regime is achieved [35]. In these cases, the different surface wettability and roughness cause a different distribution of the drops and wear phenomena of the lubricated tribopairs could lead to scuffing wear process [36]. In this framework, the combinations of variations in surfaces roughness, in some cases affected by MT phenomena [12,37] and wettability of the synovial lubricated surfaces, assumes a particularly interesting significance.
Published data on wear-rate of retrieved alumina after in vivo function have been highly variable. Boutin et al. [38], through the examination of 35 retrieved implants, found that the long-term success of alumina-alumina [39] total hip replacement depends on the ceramic microstructure (wettability and roughness), on the implant geometry, and initial positioning and stability of cup fixation. Winter et al. [40], examining one hundred COC THA, found a significant number of cases in a very unfavorable result such as extensive wear, ball fracture, or migration of the socket. Obviously, it was stressed that the majority of the cases refer to prostheses implanted prior to 1990, and are associated either to inhomogeneous granulometry or irregularity of the surface, even if the prostheses' design was similar to those used in our study.
Topographical surface characteristics were proven to improve the biological, chemical, and tribological properties of implant coupling [41]. Surface modification techniques were applied on the surface of hip prostheses in order to reduce the friction and wear of artificial implants and so improve their tribological performance. Recent investigations proved that, in some load and kinematical conditions, a full fluid film lubrication cannot manifest, and the lubrication, whether boundary or mixed, is related to the surface interaction, even if in the presence of the synovial lubricant [34,37]. In these conditions, there is a particular interest in the combination between the topographical microcharacteristic of the coupling surfaces and their wettability properties.

Conclusions
Biotribological investigations show a clear dependence between the coupling surfaces' shapes, friction, and wear phenomena acting in a generic prosthetic synovial lubricated joint. The present investigation represents a first step toward a challenge to generate a systematic correlation between wettability, surface topography and MT phenomena of total hip replacements' COC femoral heads. The experimental investigation was conducted taking into account three different prostheses groups: brand new, in vitro tested, and explanted femoral heads. The results highlight that the average roughness values, measured on the explanted specimens, were greater with respect to the new ones and the in vitro specimens. All the brand new femoral heads, with lower values of surface roughness, showed high values of wettability angle ensuring the femoral heads hydrophilic behavior in water environment and allowing the authors to discuss the results in the framework of scientific literature on the role of wettability on lubrications mode in artificial synovial joints.
Unfortunately, due to the limited amount of available data, it is still not possible to achieve a full statistic correlation of the measured data, but they could be very useful to the entire scientific community researching these topics. The potential to correlate surface topography, MT and wettability represents a challenge to overcome, for which this investigation should constitute the starting point.