Stability and trunnion wear potential in large-diameter metal-on-metal total hips: a finite element analysis

Abstract

Background: Large-diameter femoral heads for metal-on-metal THA hold theoretical advantages of joint stability and low bearing surface wear. However, recent reports have indicated an unacceptably high rate of wear-associated failure with large-diameter bearings, possibly due in part to increased wear at the trunnion interface. Thus, the deleterious consequences of using large heads may outweigh their theoretical advantages.

Questions/purposes: We investigated (1) to what extent femoral head size influenced stability in THA for several dislocation-prone motions; and the biomechanics of wear at the trunnion interface by considering the relationship between (2) wear potential and head size and (3) wear potential and other factors, including cup orientation, type of hip motion, and assembly/impaction load.

Methods: Computational simulations were executed using a previously validated nonlinear contact finite element model. Stability was determined at 36 cup orientations for five distinct dislocation challenges. Wear at the trunnion interface was calculated for three separate cup orientations subjected to gait, stooping, and sit-to-stand motions. Seven head diameters were investigated: 32 to 56 mm, in 4-mm increments.

Results: Stability improved with increased diameter, although diminishing benefit was seen for sizes of greater than 40 mm. By contrast, contact stress and computed wear at the trunnion interface all increased unabatedly with increasing head size. Increased impaction forces resulted in only small decreases in trunnion wear generation.

Conclusions: These data suggest that the theoretical advantages of large-diameter femoral heads have a limit. Diameters of greater than 40 mm demonstrated only modest improvement in terms of joint stability yet incurred substantial increase in wear potential at the trunnion.

Clinical relevance: Our model has potential to help investigators and designers of hip implants to better understand the optimization of trunnion design for long-term durability.

Fig. 1

An intraoperative image shows an MOM implant trunnion during revision THA from a patient with pain and elevated metallic ions 3 years postoperatively. Visually extensive corrosion (white arrow) was identified along the distal aspect of the trunnion. Gross macroscopic wear was identified on the trunnion surface, especially at the leading and trailing edges (blue arrows). Histology confirmed adverse tissue reaction to metal debris.

Fig. 2A–G

The finite element model consisted of THA hardware, the proximal femur, and (A) the hemipelvis, and (B) the hip capsule. (C) When appropriately seated on the tapered trunnion of the neck, a moment arm exists between the center of rotation of the head and the center of pressure on the trunnion. Femoral head diameters used in this study included (A) 32, (B) 36, (C) 40, (D) 44, (E) 48, (F) 52, and (G) 56 mm. The trunnion-bore geometry was identical for all neck-head assemblies.

Fig. 3A–C

Impaction seating of the trunnion-head interface was performed by (A) simulating a single impulsive (20-millisecond) load to the femoral head in a direction parallel to the axis of the trunnion, (B) resulting in initial contact between the trunnion and head bore. (C) The head was firmly seated at the end of the impaction impulse. Loads of 5 kN (baseline), 1 kN (low), and 17 kN (high) were considered [22, 33].

Fig. 4

A contact analysis convergence study was used to determine the appropriate mesh density used for the trunnion-bore assembly. Peak values of contact stress at the interface were recorded during the head seating analysis step. A nominal mesh density of 10 elements/mm³ was deemed to provide suitably accurate solutions, consistent with reasonable computational economy. CPU = central processing unit.

Fig. 5A–E

Femoral head subluxation was tracked during the entire input kinematic sequence for each simulation. Numerical values of subluxation during the stooping, sit-to-stand low, sit-to-stand normal, squatting, and foot pivoting dislocation challenges were averaged for each of the 36 separate cup orientations (rendered here in radiographic orientation) for a femoral stem in 10° of anteversion. (A) For the 32-mm head diameter, average femoral head subluxation was greatest for cups positioned in low values of acetabular anteversion and inclination and for cups positioned in high anteversion. Subluxation was similarly determined for head diameters of (B) 36, (C) 40, (D) 44, and (E) 48 mm.

Fig. 6

The percentage of orientations yielding a stable articulation (defined as having a maximum femoral head subluxation < 1 mm) was shown to increase as a function of femoral head diameter. The effect of increasing head size was greatest when moving from 32 mm to 36 mm, while there was progressively diminishing improvement in stability with the higher increases of head diameter.

Fig. 7A–C

Peak values of surface von Mises (vM) stress occurring at the articular surface were determined for each simulation. (A) For gait, bearing surface stress decreased precipitously when head diameter was increased from 32 mm to 36 mm, for all three cup orientations. Progressively more modest decrements in bearing surface stress with increased head diameter were observed for head sizes of more than 36 mm. Similar relationships were observed for (B) stooping and (C) sit-to-stand for the neutrally and vertically positioned cups. For both of these dislocation challenges, surface stress was greatly increased for horizontally oriented cups, due to impingement, subluxation, and significant edge loading.

Fig. 8A–C

Peak values of von Mises (vM) stress occurring at the trunnion were determined for each simulation. (A) For gait, trunnion stress was found to increase with increased femoral head diameter for all three cup orientations. The influence of head diameter on trunnion stress was progressively more pronounced for higher head diameters. A similar dependency of head diameter on trunnion stress was also observed for (B) sit-to-stand and (C) stooping. Stooping simulations with horizontally positioned cups at lower values of head diameter resulted in frank dislocation before maximum joint contact stresses could be developed, resulting in lower values of trunnion stress. For stooping, stresses at higher values of head diameter were found to approach or exceed the yield stress of cobalt-chromium (broken horizontal line), indicating potential for plastic yield on the trunnion for larger-head diameters.

Fig. 9A–B

Micromotion of the trunnion relative to the head bore results in the generation of linear wear on the interface surfaces. Total cumulative linear wear for (A) gait at the end of the motion cycle was approximately ½ of that occurring for (B) sit-to-stand. For all motions, linear wear increased with increased femoral head diameter.

Fig. 10A–C

Volumetric wear at the trunnion demonstrated strong dependence on head size for (A) gait, (B) stooping, and (C) sit-to-stand. This dependency of head size was most pronounced at higher values of head diameter.

Fig. 11

Relative nodal slippage (micromotion) during the input assembly/motion simulations at six distinct instances (i to vi) is shown. Direction of nodal slip is indicated by arrows of vectors assigned to selected nodes on the trunnion interface. (i) Trunnion assembly during impaction resulted in vertical slip of interface nodes. (ii) After the impaction load, elastic recoil in the trunnion-bore interface resulted in relaxation of the assembly and slip in the opposite direction. (iii) During the motion sequence, node slip was minimal during low values of joint contact load and bearing surface frictional torque. (iv) For intermediate values of both bearing surface frictional torque and joint contact force, net slipping was primarily rotational. (v) Further increasing both bearing surface frictional torque and joint contact loads resulted in a mixed rotational/translation slip. (vi) At peak values of joint contact load, slip was primarily parallel to the axis of the trunnion-bore.

Fig. 12

The finite element simulation consisted of three distinct phases (i to iii). (i) The femoral head was impacted onto the trunnion, resulting in the rapid development of contact pressure at the interface. (ii) After the impulse loading, the trunnion-bore interface underwent elastic recoil until static equilibrium was reached. (iii) Then, the motion cycle (gait in this illustration) was run. While the use of higher impaction loads resulted in substantially greater initial contact pressure, all three simulations converged toward similar values of pressure during the duty cycle simulation. The illustration shown is for neutrally positioned cups with 56-mm femoral head diameters.

Fig. 13

Increased interface motion, measured in terms of instantaneous slip distance between nodes at the interface, was seen for the gait simulation with a low (1-kN) impaction load, compared with normal (5-kN) or increased (17-kN) loads for 56-mm head diameters in neutral cup positions.

Fig. 14

Increasing the value of impaction force decreased the volumetric wear occurring at the trunnion interface during the gait cycle for 56-mm head diameters in neutral cup positions. This relationship exhibited exponential decay (inverse power law regression).