Long-term issues with THA are linked to bearing surface wear and particle-induced osteolysis [30]. Strategies aimed at reducing wear and osteolysis have been focused on improving the material properties, as well as developing an alternative, e.g., ceramic, bearing surface [31]. Despite the increasing use of COC bearings in THA over the past decade, concerns regarding this bearing surface, including ceramic head or liner fracture, noise generation, and malseating of the liner, remain. However, few reports have described the clinical and radiographic outcomes following modern PA and AMC COC THAs using a single cementless hip system in a large cohort of patients with mid- to longer-term follow-up. Therefore, we investigated the mid-term clinical and radiographic outcomes, including the prevalence of PJI, osteolysis, component loosening, and dislocation; analyzed the incidence of bearing-related complications, such as ceramic fracture and noise generation; and determined the existence of specific patient or surgical factors associated with noise generation using data from the KHR.
Many previous studies regarding third- and fourth-generation ceramic bearings have found little or no osteolysis [1, 4, 10, 11, 32, 33]. In the present study, we also found no evidence of acetabular or femoral osteolysis at the final follow-up. The survival results were encouraging and similar to those of previously published reports on modern COC bearings [1, 8, 10, 31,32,33,34,35]. For Forte COC bearings with alumina 28-mm heads, Kim et al. [33] reported a 99.7% 20-year overall survival rate for all-cause revision. In addition, D’Antonio et al. [31], in their multi-surgeon, multi-center study, reported 10-year survival rates of 97.9% for System I and 95.2% for System II; Lee et al. [34] reported a 10-year survival rate of 99%, and Murphy et al. [35] reported a 9-year survival rate of 96%. For the Delta COC bearings, Hamilton et al. [1] reported 5-year survival rates of 97.7% for the 28-mm COC arm and 97.3% for the 36-mm COC arm, and Kim et al. [32] reported a 13-year survival rate of 99.7%, predominantly for the 36-mm femoral head. Survival rates in our series with regard to all-cause revision were 98.4% for the Forte COC cohort and 98.6% for the Delta COC cohort at 5 years. PJI is a disastrous and challenging complication occurring after THA [5, 36]. In the present study, no hip had an early or late PJI in either the Forte or Delta COC group. Some recent reports have suggested that infection is associated with bearing surface type [5, 6, 37, 38]. Pitto et al. [5], in an analysis of New Zealand Joint Registry data over a 15-year period, suggested that COC bearings are associated with a lower risk of PJI compared with ceramic-on-polyethylene, metal-on-polyethylene, and metal-on-metal bearings, but emphasized that these findings must be considered to be extremely preliminary. The authors found no difference in the rate of early PJI in the first 6 months among bearing surfaces whereas COC bearings were associated with a lower risk of revision for PJI over the entire period of observation. Our results also support these preliminary observations. Although the whole issue of PJI is complex and multifactorial, these results are likely to be related to the material properties of the ceramics. Ceramic bearings produce lower wear and their debris is more inert than that arising from any other bearing surface [35, 39,40,41,42]. Thus, a smaller amount of periprosthetic debris and subsequently reduced local tissue reaction can result in a protective effect against infection after THA [43].
The risk of dislocation after THA is also related to the bearing couples, although the cause of dislocation is too multifactorial [3, 4]. Hernigou et al. [3] found that, with no difference in cup orientation, COC bearings decreased the risk of dislocation, especially late dislocation, compared with ceramic-on-polyethylene bearings. They suggested that the reasons for the lower rate of dislocation with COC bearings were attributable to the more fibrous, thick capsule and reduced fatty atrophy in the periarticular muscles observed at the time of the revision, and that these differences might occur as a result of different biological responses to wear byproducts generated by different bearing surfaces [3, 4]. Other authors similarly reported lower dislocation rates in THA using Delta ceramics, ranging from 0.5 to 1.1% [10, 44, 45]. We found a lower dislocation rate (0.3%) in the Delta COC cohort, treated predominantly with 36-mm heads, than in the Forte COC cohort (1.9%), although this difference was not significant (p = 0.124). This finding is consistent with results reported in the literature (3.4% for the 28-mm COC vs. 1.8% for the 36-mm COC) [1]. The use of a larger femoral head with Delta ceramics combined with the lower wear rate would provide a major advantage, including improved stability, reduced impingement, and, consequently, a lower dislocation rate [8].
Despite advances in the manufacture of modern ceramics, ceramic material fracture remains a major concern. Recent results reported in the literature have demonstrated that the Delta ceramic is more fracture resistant than the Forte ceramic and, thus, that it has significantly reduced the risk of fracture [1, 7, 46, 47]. Massin et al. [46] performed a systematic review of the literature and reported head fracture rates of 0–10% with the BIOLOX® Forte ceramic, with a median close to zero. According to the French ceramic experience, the rates of head fracture were 0.18% with the Forte ceramic and 0.0013% with the Delta ceramic, which represent a 100-fold difference in favor of the Delta ceramic, whereas the manufacturer’s data (CeramTec) revealed a 20-fold difference (0.021% with Forte vs. 0.001% with Delta) [7, 46]. In contrast, the fracture rate of liners has remained stable at approximately 0.03 to 0.08%. We observed one (0.3%) fracture of a 28-mm short-neck alumina Forte head at 4.5 years after the index THA, and no ceramic fracture in the Delta COC cohort. Our findings are consistent with those of Kim et al. [32], who identified no Delta ceramic head or liner fracture at a mean of 13.1 years of follow-up. Ceramic fractures are usually associated with specific events, such as trauma, hip dislocation, or cup malposition (i.e., excessively abducted or anteverted cup angle) [7]. The patient in the current study experienced painful squeaking during her usual activities from 21 months after the index surgery and subsequently had a traumatic event, although the cup orientation was within the normal range. Abdel et al. [48] also found that painful squeaking following COC THA in four patients was related to ceramic liner fractures. They emphasized the need for a more thorough investigation, as squeaking was associated with increased pain.
Noise, especially squeaking, after THA has become a particular issue with ceramic bearings [11, 49]. Overall, the prevalence of squeaking has been reported in the English literature to range from 2 to 21% [11, 13, 48]; Stanat et al. [12] determined an average incidence of 2.4% in their meta-analysis. In the current study, the overall incidence of noise of any type was 7%, with the most common type of noise being clicking. Squeaking was reported in 2% of hips in each group, with no hip being revised due to this phenomenon. Although our results are similar to those of previous reports [11, 32] on modern ceramic bearings, the 2% incidence of squeaking in our study was lower than the rate of 7.5% published recently by Hamilton et al. [1]. The reason for this difference is unclear, but may be due to the influence of lower BMI in our series compared with Westerners. Although the exact etiology of squeaking remains uncertain, cup malposition, edge loading of components, decreased lubrication between the bearing surfaces, and metal transfer to the femoral head have been implicated as contributing factors [11]. In a meta-analysis, Stanat et al. [12] found BMI to be the only associated patient factor. Walter et al. [49] reported that cup malposition was associated with squeaking. In contrast, other authors [11, 13] did not find this association. These results are consistent with our findings, and we were unable to identify any factor that significantly affected noise generation.
This study has several limitations. First, despite the relatively large cohort compared with previously published studies [8, 10, 31, 34, 35, 45], our sample was not large enough for subgroup analyses of factors such as ceramic fracture or noise generation. In addition, the utilization of ceramic head size could not be matched between groups and the two cohorts were not comparable in the ceramic head size. In accordance with the manufacturer’s guidelines, 28 mm and 32 mm heads were only available for Forte ceramic bearings, and 32 mm and 36 mm heads for Delta ceramic bearings. The utilization of smaller heads in the Forte COC cohort can be a cause of concern regarding ceramic fracture (0.3% vs. 0%, p = 1.000) and dislocation (1.9% vs. 0.3%, p = 0.124) although not significantly different in this study. Second, the study was conducted using a single institution’s series from the registry database because data utilization for multicenter studies is not yet available in the KHR. However, we believe that the homogeneity of the current study, including the use of the same implant, surgeon, and surgical approach, can mitigate this shortcoming. Moreover, during the study period, we used only COC bearings for all primary THAs, regardless of patient age; therefore, selection bias may have been avoided. Third, although cup abduction and anteversion angle were measured, stem anteversion was not assessed. Moreover, computed tomography was not performed to evaluate osteolysis. However, our findings from standardized plain radiographs, which were assessed rigorously according to the criteria of Engh et al. [24], were similar to other published results. Fourth, we did not evaluate patient-reported outcome measures or activity levels or the influence that audible sounds had on them. Finally, this study included only East Asians with a mean BMI of 23.4 kg/m2. Therefore, these findings may not be generalizable to a Western population or to highly obese patients.