A comparison of metal/metal and ceramic/metal taper‐trunnion modular connections in explanted total hip replacements

Abstract Corrosion and wear are commonly found at the taper‐trunnion connection of modular total hip arthroplasty (THA) explanted devices. While metal/metal (M/M) modular taper‐trunnion connections exhibit more wear/corrosion than ceramic/metal (C/M) modular taper‐trunnion connections, damage is present in both, regardless of material. This study used a combination of assessment techniques including clinical data, visual scoring assessment, optical imaging, profilometry, and x‐ray photoelectron microscopy (XPS), to investigate wear mechanisms and damage features at the modular taper‐trunnion connection of 10 M/M and 8 C/M explanted THAs. No correlation was found between any demographic variable and corrosion wear and assessment scores. All assessment techniques demonstrated that the stem trunnions had more damage than head tapers for both explant groups and agreed that C/M explants had less corrosion and wear compared to M/M explants. However, visual assessment scores differed between assessment techniques when evaluating the tapers and trunnions within the two groups. Profilometry showed an increase (p <.05) in surface roughness for stem trunnions compared to head tapers for both explant groups. X‐ray photoelectron spectroscopy performed on deposits from two M/M explants found chromium and molybdenum carbides beneath the surface while chromium sulfate and aged bone mineral were found on the surface suggesting that the debris is a result of corrosion rather than wear. These results indicate that taper‐trunnion damage is more prevalent for M/M explants, but C/M explants are still susceptible to damage. More comprehensive analysis of damage is necessary to better understand the origins of taper‐trunnion damage.


| INTRODUCTION
Biological issues with wear performance and response to metal wear debris led the Food and Drug Administration to reclassify total hip arthroplasty (THA) implants with metal-on-metal bearings as Class III medical devices 1 based on the recommendation from an Orthopaedic and Rehabilitation Devices Panel. While there are currently no FDAapproved metal-on-metal THAs marketed for use in the United States, the use of metal heads on metal stems that creates a metal/metal (M/M) taper-trunnion modular connection still allows for wear/corrosion at the modular interface which has led to clinical reports of adverse local tissue reactions (ALTRs) in THAs without metal-onmetal bearings. [2][3][4][5] Ceramic heads, made either from alumina or zirconia-toughened aluminum, on metal stems create a ceramic/metal (C/M) taper-trunnion modular connection and have been shown to reduce the amount of wear and corrosion at the modular taper connection 6,7 resulting in their use as an alternative to metal heads. 8,9 The potential for wear/corrosion between the ceramic head and the metal stem remains due to the difference in hardness between metals and ceramics, and thus further characterization is still needed.
Damage of modular head tapers has been assessed using a number of metrics including visual assessment, [10][11][12] volume loss measurements, 6,12 mass loss measurements, 13,14 as well as electrochemical assessment of the contacting surfaces. 13,14 Visual assessments are generally qualitative and are based on the subjective judgment of multiple observers who rate the visual or photographic surfaces of the modular connections and provide a score based on the relative quantity and intensity of the observed wear/corrosion on the surfaces. [10][11][12] However, scoring methods with different scales and scoring criteria can make it difficult to compare results across different studies. Volume loss measurements are more quantitative than the visual assessment and use coordinate measurement machines to measure changes in the taper/trunnion volume compared to the ideal or as-manufactured volume. 6,12 Mass loss is a quantitative assessment where components are weighed before and after testing to measure the amount of material removal from the contacting surface. 13,14 Measuring the electrochemical potential while the modular surfaces are dynamically loaded allows for a direct measurement of corrosion and re-passivation of the modular surfaces. 13,14 Of the four techniques, visual assessment and volume loss are most suited for assessing explanted modular head tapers since lack of preimplantation mass data and tissue fixation can prevent accurate mass measurements of the explanted devices and it is not possible to perform electrochemical measurements of the modular head tapers once explanted. While many of these previous studies have shown a correlation between the different metrics, the need for multiple metrics underscores the complexity of assessing the performance of modular connections.
Previous studies have provided significant insight into the performance of modular taper-trunnion connections in THA and there have been a number of studies characterizing M/M taper-trunnion performance. A study investigating titanium stems coupled with titanium adapter sleeves showed that explants with severe head taper corrosion had greater bearing wear and elevated chromium and cobalt serum ion levels. 15 A study on explanted cobalt chromium heads on metal stems showed no difference in fretting and corrosion as a function of head taper size. 16 Fretting corrosion comparisons between cobalt chromium heads and titanium or cobalt chromium stems showed significantly more material loss in cobalt chromium stems. 17 For ceramic heads, previous research has shown that mechanically assisted crevice corrosion occurs to a lesser extent, and the wear rate is an order of magnitude lower: 0.01 mm 3 /10 6 cycles for ceramic head tapers with metal stem trunnions versus 0.1 mm 3 /10 6 cycles for M/M taper-trunnion devices. 18 This has been attributed to better mechanical interlocking of contact surfaces, which reduced micromotion, as well as the ceramic heads having greater hardness than their metal counterparts. However, due to the paucity of studies involving ceramic head tapers with metal stem trunnions, these findings cannot necessarily be generalized to more than one brand of hip implant, nor can significant connections be drawn between ceramic head taper corrosion and patient-related factors, such as implantation time. 7,9 The goal of this study was twofold: (a) to investigate the differ-   Femoral head tapers and stem trunnions were scored independently so that observers did not know which components had been implanted together. The extent of corrosion on the engaged portion of the taper and trunnion surface was assessed by visually inspecting the digital images and scored using two different scoring methods, the

5-point Anderson Criteria for Corrosion 19 and the 4-point Goldberg
Criteria for Corrosion and Fretting, 10 as presented in Di Prima et al. 11 If there was a discrepancy in scores between different observers for a particular component, the median score of the three observers was used for subsequent analyses.

| Digital mosaic method
The entire contact surface of each femoral stem trunnion and head taper component was imaged using custom fixtures to maintain relative positioning between images. Using the Digital Mosaic Method (DMM), 11 individual image sections were digitally stitched into their respective detailed mosaic of the entire contact surface for all components. Three independent observers identified and manually drew contour boundaries around regions of wear (defined as where the machining marks were damaged or no longer visible) and corrosion (visible grain boundary corrosion or build-up over the machining marks) using the ImageJ software (NIH, Bethesda, MD). The areas of the identified damage were then added (thereby combining wear and corrosion) to determine the total damaged area, and normalized by dividing by the total contact area for a possible score of 0-100. All images were randomized to ensure that there was no observer bias caused by the assessment of the matching surface.

| Profilometry
Surface roughness (R a ) evaluations were made for all surfaces using a noncontact optical profilometer (ContourGT-1, Bruker Corp., Billerica, MA). All measurements were taken using white light and 10Â objective and 0.55Â magnifier resulting in a measurement area of 1. 16     The damage features observed can be categorized into four general groups: fretting, chemical dissolution of surface oxides (dissolution), intergranular corrosion, and surface striations ( Figure 2). In general, there was more evidence of wear mechanisms and corrosion on the trunnions than on the tapers. Although the M/M explants exhibited more evidence of wear mechanisms and corrosion, the C/M explants also showed evidence of wear, corrosion, and metal transfer.

| Statistical analysis
The wear mechanisms and corrosion were visually different between the two types of explants ( Figure 3). All   nents that had evidence of wear and/or corrosion, two or more types of damage features (wear, fretting, surface striations, dissolution, and intergranular corrosion) were commonly observed. Figure 3 and Table 3

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.