Ceramic–Ceramic Hip Implants: Investigation of Various Factors Associated with Squeaking

Abstract

Despite the low wear rate of ceramic–ceramic hip implants, hard-on-soft bearings remain the most commonly used bearings in North America and Western Europe. A major concern with ceramic–ceramic hip implants is the occurrence of squeaking phenomena, which are still not fully understood. Various factors are mentioned in the literature, but currently, studies mostly focus on only one specific parameter. The goal of this study was to systematically analyze four different factors (cup orientation, protein concentration of the test fluid, contact pressure and head roughness) that may influence the squeaking behavior of this bearing type. An in vitro simulation according to ISO 14242-1 was performed using an AMTI Vivo simulator, and acoustic signals were recorded. No squeaking occurred for any of the four parameters when bovine serum or water was used as the test fluid. Squeaking was observed only under dry conditions with the ceramic–ceramic bearing. Under dry conditions, the maximum resulting torque increased steadily, and squeaking occurred after approximately 300 cycles at a resulting torque of more than 22 Nm. Thus, the resulting torque might be one factor leading to squeaking and should be kept low to reduce the risk of squeaking.

1. Introduction

Using a ceramic–ceramic bearing in hip implants leads to the lowest amount of wear compared to other material combinations that are clinically used [1,2]. Nevertheless, only a small percentage of primary hip implantations in Western Europe and North America are performed with ceramic–ceramic bearings [3,4,5]. One issue with ceramic–ceramic hip implants is the potential for squeaking phenomena, which occur in 4.2% of ceramic–ceramic hips and may be present during various types of motion for the patient, such as walking, stair climbing or squatting [6,7]. Patients with a squeaking ceramic hip typically do not experience pain or reduced function in their artificial joint and therefore do not necessarily require revision surgery [8,9]. However, patients with squeaking ceramic hips showed significant lower satisfaction compared to those with non-squeaking ceramic hips [10]. The reasons for squeaking phenomena are not yet fully understood. Owen et al. named increased friction, acetabular orientation, microseparation, femoral stem design, rim impingement and patient-specific factors as possible contributors to squeaking ceramic hips [6]. In general, suboptimal lubrication, like a breakage of the fluid film, is associated with squeaking phenomena, which may be caused by the previously mentioned factors or metal particles in the ceramic–ceramic bearing [11,12]. In clinical studies, squeaking hips can be analyzed directly while the affected person undergoes different types of motion. However, due to patient-specific differences, it is challenging to identify a single general factor that leads to squeaking. This is reflected in the differing results of clinical studies on factors such as the influence of cup orientation on squeaking [7,9,13]. In vitro simulation has the potential to isolate specific factors that may cause squeaking phenomena by keeping all other factors constant. Previous studies have typically focused on one or at most two factors which may cause squeaking or used a simplified in vitro test setup [14,15,16,17]. Such approaches help identify possible factors that may lead to squeaking. However, due to the different test setups and ceramic components used in these experimental studies, the generalization of factors leading to squeaking is also barely possible. To determine the impact of different factors systematically, an analysis using the same test setup would be desirable. Therefore, in the current study, four different factors and their impact on squeaking are analyzed using one single in vitro setup, aiming to provide a better understanding of each factor’s influence on squeaking in ceramic–ceramic bearings. The following research questions will be addressed within this study:

(1)

Does a variation in inclination and anteversion lead to squeaking?

(2)

Does a variation in the protein concentration of the test fluid, or even dry test conditions, lead to squeaking?

(3)

Does an increased contact pressure lead to squeaking?

(4)

Does an increased surface roughness lead to squeaking?

2. Materials and Methods

2.1. Test Specimen

In total, five BIOLOX^®delta ceramic femoral heads (3× size 36, 1× size 34, 1× size 28) (CeramTec, Plochingen, Germany) and two BIOLOX^®delta ceramic acetabular inserts of size 36 were used for this study. In

Table 1. Implant components used for each research question.

Research Question	Femoral Head Number	Femoral Head Size	Rough or Polished Femoral Head	Insert Number	Insert Size
1	#1	36 mm	polished	#1	36 mm
2	#2	36 mm	polished	#1	36 mm
3	#4 and #5	28 and 34 mm	polished	#2	36 mm
4	#3	36 mm	roughened	#2	36 mm

, the femoral heads and inserts are assigned to the four research questions.

As shown in Table 1, the ceramic components were partly reused for different research questions. If squeaking occurred, the implants were analyzed for defects. If defects were visible, the damaged component was replaced by a new implant component. The initial condition of the ceramic components was polished, as received by the manufacturer. For research question 4, the femoral head was roughened by a shot peening process with a peening pressure of 6 bar using corundum grains with sizes between 0.60 and 0.85 mm.

2.2. Test Setup

For in vitro testing, a VIVO 6-degrees-of-freedom (6-DoF) joint motion simulator (AMTI, Watertown, MA, USA) was used. Adapters were designed and machined to fixate the femoral head and the ceramic insert in the test machine. To address research question 1, a variation in the insert inclination and anteversion was necessary. Therefore, wedges with different angles were used, which could be attached perpendicular to each other and placed between the VIVO simulator and the ceramic insert (see Figure 1).

The test chamber, consisting of the femoral head, insert and its adapters, was sealed using a plastic tube to allow for lubrication of the ceramic bearing. Two heating elements were included in the test setup, together with a temperature control unit to keep the test fluid at a uniform temperature of 37 ± 1 °C. To record squeaking phenomena during the in vitro simulation, a camera (GoPro HERO7, GoPro, Inc., San Mateo, CA, USA) was set up in front of the simulator, and a microphone (RØDE XDM-100, RØDE Microphones, Silverwater, Sydney, Australia) was placed next to the test chamber. The microphone was connected to the camera via an aux cable to synchronize the video and audio signals. The complete setup is shown in Figure 2.

2.3. In Vitro Simulation

For each research question, motion and load profiles were applied according to ISO 14242-1 with a flexion–extension motion from +25° to −18°, an abduction–adduction motion from −4° to +7° and an internal–external rotation from +2° to −0°. The duration of each test was 1000 cycles at a frequency of 1 Hz. Before every test, the chamber was filled with 150 mL of test fluid and heated to 37 ± 1 °C.

2.3.1. Parameters for Research Question 1

To detect possible squeaking phenomena at different cup angles, 25 different combinations of inclination and anteversion were tested. According to Murray, the radiographic inclination angle is defined as the angle between the longitudinal axis of the body and the projection of the acetabluar axis in the coronal plane, and the anteversion is defined by the angle between the acetabular axis and the coronal plane [19]. The inclination angles were 20°, 30°, 40°, 50° and 60°, and the anteversion angles were 0°, 5°, 15°, 25° and 35°. Bovine calf serum with a protein concentration of 30 ± 1 g/L was used as a test fluid. No additives like sodium azide (NaN₃) and ethylene diamine tetraacide (EDTA) were added because only short-term tests were performed.

2.3.2. Parameters for Research Question 2

To analyze the impact of the test fluid, different fluids were used. After using the bovine calf serum with a protein concentration of 30 g/L, bovine calf serum with protein concentrations of 73 g/L (raw serum), 20 g/L, 10 g/L and 0 g/L (demineralized water) was used. In addition, a test without any test fluid (dry conditions) was conducted. The cup was positioned at an inclination of 40° and an anteversion of 0°.

2.3.3. Parameters for Research Question 3

To determine possible squeaking phenomena at higher contact pressures, differently sized ceramic femoral heads were used. In addition to the 36 mm femoral head, a 34 mm femoral head and a 28 mm femoral head were combined with a 36 mm ceramic insert. The pressure, according to the contact theory of Hertz, at the maximum load of 3000 N while assuming a radial clearance of 40 µm for the 36 mm femoral head can be calculated. The 36 mm femoral head leads to a maximum contact pressure of approximately 68 MPa; the 34 mm femoral head leads to a maximum contact pressure of about 617 MPa; and the 28 mm femoral head leads to a maximum contact pressure of approximately 1736 MPa. The tests were performed in bovine calf serum with a protein concentration of 30 g/L, and the cup was positioned at an inclination of 40° and an anteversion of 0°.

2.3.4. Parameters for Research Question 4

To analyze the influence of increased surface roughness on squeaking, a standard polished inlay was combined with a roughened (shot peened) ceramic femoral head. The roughness of the roughened femoral head was measured according to ISO 4287 [20], determining the surface roughness, R_a (0.633 ± 0.047) and R_z (4.457 ± 0.384), using a tactile roughness measurement device (Perthometer M2, Mahr, Goettingen, Germany). The same roughness measurements were performed for the polished head, with an R_a of 0.022 ± 0.002 and an R_z of 0.027 ± 0.003. The roughness measurements were performed three times at spots not in contact with the ceramic insert during in vitro testing. The tests were performed in bovine calf serum with a protein concentration of 30 g/L. The cup was positioned at an inclination of 40° and an anteversion of 0°.

2.4. Data Collection

For each test, videos were recorded, and acoustic signals between 20 and 20.000 Hz were captured. If squeaking phenomena occurred, a fast Fourier transform (FFT) function was applied to the acoustic signals using a self-made MATLAB script (The MathWorks Inc., MATLAB R2023a, Natick, MA, USA). In cases of squeaking, a reference measurement, where no squeaking occurred, was used to identify and account for the systematic sounds from the test setup. The reference measurement revealed that the in vitro simulation without squeaking noises led to sounds with frequencies below 1 kHz. Thus, the frequency of squeaking sounds could be detected because these sounds have frequencies higher than 1 kHz [17,21].

In addition, the torques occurring during the in vitro simulations were recorded with a sampling rate of 100 values per cycle. For each test, 10 cycles (cycles 31–cycle 40) were used to determine the resulting torque. The maximum resulting torque for each of these ten cycles was determined. The mean and standard deviation were calculated for each test to compare the different test conditions.

3. Results

3.1. Component Orientation

For all 25 component orientations, no squeaking occurred when applying the ISO load and motion profile with bovine calf serum with a protein content of 30 g/L. The averaged torque values of the 25 component orientations are shown in Figure 3.

A cup inclination of 30° in combination with an anteversion of 5° led to the lowest torque values of all the combinations.

3.2. Test Fluids and Dry Conditions

Squeaking only occurred under dry conditions when the ISO load and motion profile with a standard implant orientation of 40° inclination and 0° anteversion was applied. It occurred after approximately 300 cycles. The detected frequencies of the reference sound and the squeaking sound after applying a FFT are shown in Figure 4.

The squeaking sound was very faint and had a frequency of approximately 5 kHz, as seen in Figure 4 (right image). Frequencies below 1 kHz were induced by the simulator itself. The maximum resulting torques depending on the different lubricants are shown in Figure 5. The absolute values can be found in the Supplementary Materials.

The dry conditions resulted in the highest maximum resulting torque compared to the lubricated conditions. Water produced the lowest resulting torques of all the tested conditions. When using serum with protein contents of 10 g/L, 20 g/L and 30 g/L, an increase in the protein concentration led to a reduction in the maximum resulting torque. However, using pure serum with a protein content of 73 g/L resulted in higher torques compared to serums with lower protein contents. For the dry conditions, an increase in torque was seen until squeaking began after approximately 300 cycles. In Figure 6, an increase in torque after intervals of 100 cycles is shown until squeaking occurred.

The averaged maximum resulting torque during squeaking (between cycle 331 and cycle 340) was 22.7 ± 0.6 Nm. The absolute values of the other measured intervals in the dry conditions can be found in the Supplementary Materials. No surface defects were seen on the ceramic components after the occurrence of squeaking.

3.3. Component Mismatch

For both mismatch groups (34 mm femoral head and 28 mm femoral head in a 36 mm ceramic insert), no squeaking occurred when applying the ISO load and motion profile with bovine calf serum with a protein content of 30 g/L. The resulting maximum torques of the two mismatch groups are shown in comparison to the non-mismatch group in Figure 7.

3.4. Roughened Femoral Heads

When applying the ISO load and motion profile and testing with bovine calf serum with a protein content of 30 g/L, no squeaking occurred for the roughened femoral head. However, the resulting maximum torque of the roughened femoral head (15.80 ± 0.69 Nm) was much higher compared to the maximum resulting torque of the polished femoral head (6.45 ± 0.35 Nm), as seen in Figure 8.

4. Discussion

This study attempted to isolate different factors that may lead to squeaking phenomena in the ceramic–ceramic bearing of hip implants using a well-defined test setup. The investigated factors included cup orientation, test fluid, mismatch between femoral head and insert and the surface roughness of the femoral head. Each factor will be discussed successively based on the results of this study and the relevant literature.

According to the findings of the current study, the anteversion (0–35°) and inclination (20–60°) of the cup have no impact on squeaking when loads and kinematics are applied according to ISO 14242-1 and the tests are performed in bovine calf serum with a protein content of 30 g/L. In vivo studies have shown differences regarding the relationship between squeaking and cup orientation. Walter et al. showed that squeaking during walking is more likely with high cup anteversion [13]. Inagaki et al. reported that smaller cup anteversions are associated with noise development when comparing noisy and silent ceramic–ceramic hips [8]. Other in vivo studies did not find any correlation between cup orientation and squeaking in ceramic–ceramic hips [21,22]. Affatato et al. tested different inclination angles of the cup and their effect on squeaking in an in vitro test according to ISO 14242 [16]. Neither a cup inclination of 23° nor a cup inclination of 45° led to squeaking sounds. However, a cup inclination of 63° resulted in squeaking. In the current study, the maximum inclination tested was 60° due to the limited range of motion of the setup. Therefore, squeaking may occur at higher inclination angles. However, Affatato et al. used the older Biolox forte rather than Biolox delta ceramic and a 28 mm ceramic coupling instead of a 36 mm ceramic bearing as used in the current study. In addition, different simulators were used in the study by Affatato et al. and the current study. In the current study, a relationship was observed between cup orientation and the maximum resulting torques. An inclination of 30° in combination with an anteversion of the cup between 5° and 15° produced the smallest maximum resulting torques. Bishop et al. also found increased friction at a high inclination of 60° compared to lower inclination angles [23]. However, they tested metal–metal bearings, which are not comparable to the ceramic–ceramic bearing used in this study.

In the current study, no squeaking occurred when applying the loads according to ISO 14242-1 as long as lubricant was present in the ceramic–ceramic bearing. Light squeaking sounds were recorded with a frequency of approximately 5 kHz when testing under dry conditions. Other studies also found squeaking sounds in a similar frequency range when testing ceramic–ceramic bearings in dry conditions. Rodgers et al. found squeaking sounds between 2 and 5 kHz when testing ceramic–ceramic retrievals [24]. O’Dwyer Lancaster-Jones and Reddiough found squeaking sounds of between 2 and 8 kHz when applying ISO 14242-1 loads under dry test conditions [17]. In the current study, the highest maximum resulting torques were found when applying dry conditions as soon as squeaking occurred (after approximately 300 cycles). Comparing the different lubricants, water led to the lowest and pure serum to the highest resulting torques. Similar results were found by Brockett et al. when testing ceramic–ceramic hips with different serum protein concentrations [25]. Pure serum led to the highest and water to the lowest frictional torques.

According to the results of this study, a component mismatch between the ceramic femoral head and the ceramic insert did not lead to squeaking when applying the ISO 14242-1 loads with bovine calf serum as a lubricant, despite the high maximum contact pressures. To the authors’ knowledge, no previous in vitro study has investigated squeaking phenomena in ceramic–ceramic bearings with a bearing mismatch of 2 mm or 8 mm. Brockett et al. analyzed squeaking effects at different clearances in metal–metal hip implants [25]. They found the highest torques and the highest incidence of squeaking in the highest clearance group. In the current study, a decreasing maximum resulting torque was found with increasing mismatch. However, the differences between the torques in the three groups were very minimal. The results cannot be compared directly to those of Brockett et al. due to the different bearing materials that were used. Hothan et al. tested ceramic–ceramic hip implants with different clearances and loads in an in vitro test setup to analyze squeaking phenomena [26]. They found that clearance had no effect on squeaking but observed that higher loads increased squeaking frequency. However, these results cannot be compared to the current study due to differences in the lubrication conditions.

In the current study, no squeaking occurred when using a roughened femoral head under ISO 14242-1 load conditions with bovine calf serum as a lubricant. However, higher maximum resulting torques were found for the roughened femoral head compared to a polished femoral head. Chevillotte et al. analyzed nine retrieved ceramic–ceramic bearings that had caused squeaking in patients [27]. The retrievals were used for an in vitro test in a custom-made hip simulator. The retrievals showed a comparable surface roughness (0.6 µm) to the roughened femoral heads used in this study. Squeaking could be generated for all retrievals under dry conditions but not when lubrication was used. However, it is known that some ceramic hips lead to squeaking sounds in vivo despite lubrication. Therefore, specific in vitro tests were needed to simulate the in vivo conditions of patients with squeaking hips. If squeaking cannot be reproduced, third-body particles might be responsible for the squeaking sounds in vivo.

Some limitations need to be mentioned to better interpret the results of the current study. The loads and motions assessed in this study were restricted to those defined by ISO 14242-1. These loads and motions vary from patient to patient in vivo, so that one or more of the analyzed factors may lead to squeaking in a specific patient. However, all different factors and combinations cannot be taken into account using an in-vitro test. Therefore, the ISO standard is suitable for performing standardized motions with standardized loads. Edge-loading, impingement, third-body metal particles and their influence on squeaking were not analyzed in this study. The influence of edge-loading on squeaking effects was successfully analyzed by O’Dwyer Lancaster-Jones and Reddiough [17]. The influence of metal particles in ceramic–ceramic bearings, which can be released from the implant due to impingement, and the resulting squeaking effects have also been demonstrated by previous studies [16,28]. Next, the absolute torque values measured with the Vivo simulator should be interpreted with caution, as no external crosstalk compensation was used. The accuracy of the torque measurements taken with the Vivo simulator needs to be determined in a future study. However, the torques measured in this study can still be compared to determine the impacts of the different factors analyzed. Among other things, squeaking sounds depend on the hip stem geometry [26]. Vibration in the ceramic bearing may occur without leading to audible squeaking sounds when using a setup with another hip stem. Therefore, accelerometric measurements would be a good addition for future studies. Another limitation is that only one test was performed for each condition. Using different ceramic bearings as well as repeated measurements may lead to other results that differ from those in the current study.

The current study shows that squeaking does not occur due to one of the analyzed factors alone. It might be a combination of the analyzed factors, as well as edge-loading, impingement and/or third-body particles, that is responsible for squeaking in ceramic–ceramic hips.

5. Conclusions

In the current study, no squeaking sounds occurred across the different cup orientations, test fluids, component mismatches or surface roughnesses of the femoral component when applying ISO 14242-1 loads with lubrication. However, the ceramic–ceramic bearing led to squeaking under dry conditions after approximately 300 cycles and a steady increase in the maximum resulting torque. Thus, keeping the torques in a ceramic–ceramic bearing as low as possible might reduce the risk of squeaking.