While the double socket technique with conventional PE liners has shown low loosening rates (0–6%), it has been linked with high dislocation rates of up to 22% [1, 11, 12]. In order to provide additional stability, constrained liners can be used in combination with this technique [3, 13]. Due to their additional constraint and reduced range of motion (ROM), however, mechanical stress is increased which may in turn lead to early loosening. This is expressed in high aseptic failure rates (15–18%) mainly attributable to the dissociation of the construct at the constrained acetabular component/ cement interface. In addition, although reduced when compared with conventional PE liners, the risk of dislocation has not been eliminated by the use of constrained liners (0–18%). Even more modern types of constrained liners with focal areas of constraint with liner cut outs promising a higher range of motion have not convincingly reduced failure rates [14, 15].
Possible clinical advantages of the use of DMCs in primary and revision surgery may thus also be applied to the double socket technique. Due to the increased head size the jumping distance until the occurrence of dislocation is increased in comparison with conventional head sizes. At the same time the mobile liner articulating both with the acetabular shell and the femoral head increases the range of motion eventually limiting the edge force which in contrast may provoke early mechanical loosening in constrained liners [16].
The advantage of these common DMC characteristics was confirmed in two recent systematic literature reviews, where the application in THA revision was evaluated. Reina et al. [17] found a (re-)dislocation rate for DMCs in revision of only 2.2% when compared to 7.1% in the control group (conventional head sizes up to 40mm) after a FU duration of 4.1 years, while Darrith et al. [9] observed a 2.2% dislocation rate along with a 96.6% survivorship after 5.4 years of FU. While there has been an increased use of DMCs in both primary and revision THA there is still a lack of long-term data.
A first biomechanical evaluation of the double-socket technique in combination with DMCs was performed by Wegrzyn et al. Compared with conventional PE liners, the pull-out forces until failure of the construct were higher when using a cemented DMC. Analysis of the failure mechanism showed that the loosening occurred in the interface between the cement and the cementless metal cup, while the interface between the DMC and the cement remained intact. These biomechanical results emphasize two main aspects. Regarding the well-integrated cementless cup, a maximal roughness of the inner surface should be achieved. Therefore, all possible screws should be removed and mechanical roughening with a burr should be performed in order to promote cement indentation. Studies have shown a 20% increase in stability after shell roughening [10, 18, 19].
Regarding the DMC, an implant explicitly used for cemented implantation with peripheral radial and concentric circumferential grooves which oppose torsion and lever loading [10] should be used. Plummer et al. evaluated a mixed collective of patients undergoing revision THA with a certain number of patients treated with a DMC cemented into a well-fixed shell. While two cases of a DMC liner cementation of a modular DMC system failed, all DMCs specifically designed for cemented fixation survived [20].
Up until now, three clinical studies have been published on this topic [5–7]. Chalmers et al. [7] investigated 18 patients undergoing revision THA with a monoblock DMC construct for different reasons of which an undisclosed number had a DMC cemented into an existing well-fixed cup while the other patients underwent a full cup revision. Overall, the authors found no dissociations but three re-dislocations after a mean FU of 3 years of which one required a revision with a constrained liner while two required open reduction of the DMC. Moreta et al. [6], reporting on 10 DMCs cemented into well fixed shells, had one recurrent dislocation which was treated with resection arthroplasty due to the patient’s various comorbidities. The most recent study on this topic by Wegrzyn et al. [5] found no dislocation or dissociation of the DMC construct after a mean FU of 3.5 years in a study collective of 28 patients.
Overall, all studies had comparable baseline patient characteristics as well as mean FU durations. Furthermore, complication rates were comparable. The main contrast to the other studies is the inclusion of non-elective cases including periprosthetic Vancouver 2B and 2C fractures which made up almost 40% of our collective. The other studies only included elective patients with liner failure or instability. During FU, all three studies showed HHS scores between 71 and 88. The fact that we also included periprosthetic fractures for the “cup in cup” technique led to additional extensive surgery on the femoral side which in many cases may possibly explain the reduced functional outcome and the higher level of mortality during FU (Table 4).
Table 4
Comparison of clinical studies on the “cup-in-cup” technique with DMCs.
study | n | Age (∅) | FU (y) | comorbidities | function | dislocation (%) | dissociation (%) | time (min) | other |
Chalmers [7] | 18 | 64 | 3.0 | / | HHS 82 | 3/18 (17%) | 0% | / | 2 PPF |
Moreta [6] | 10 | 79 | 3.5 | CCI 4.3 | HHS 71 | 10% | 0% | / | / |
Wegrzyn [5] | 28 | 82 | 3.5 | 56/44% ASA 3/2 | HHS 88 | 0 | 0 | 107’ | / |
Bellova | 33 | 79 | 2.5 | ASA 2.72 CCI 5.1 | HHS 59 WOMAC 60 | 2/26 (7.7%) | 1/24 (4.2%) | 124’ | 1 infection 1 PPF |
ASA, American Society of Anesthesiologists (score); CCI, Charlson Comorbidity Index; FU, follow-up; HHS, Harris Hip Score, min, minutes; n, number; PPF, periprosthetic fracture; WOMAC, Western Ontario and McMasters University Osteoarthritis Index; y, years; ∅, average. |
An important advantage of the double socket technique is the reduced surgery duration in comparison with a full acetabular component exchange. The fact that the mean surgery duration of 124 minutes (min.) in the present study was somewhat longer compared to that reported by Wegrzyn et al. [5] (107 min.) may be attributed to the fact that some of our cases were not isolated cup revisions, but in approximately 40% of the cases combined cup and stem revisions, mostly due to periprosthetic fracture. In isolated cup revisions within our collective, the mean surgery duration was 92 min. when compared with 169 min. for combined cup and stem revisions.
We observed one cup dissociation in our study collective shortly after surgery while none were reported in the other clinical studies. The difficulty of the double socket technique is achieving firm fixation of the inner cup. Failure can occur on the side of the fixed implant due to missing surface roughness or due to an improper cement mantle possibly providing inadequate fixation of the DMC. Wegrzyn et al. recommend a cement mantle of 2–3 mm around the DMC requiring a DMC which is – depending on the thickness of the cementless cup – at least 10 mm smaller concerning the outer implant diameter of the fixed cup. If this recommendation is considered, cementless cups smaller than 54 mm would disqualify for the double socket technique. Since the size of cementless cups is often smaller than size 54, we accepted an eccentric positioning of DMC in several cases (65%) possibly with a thinner cement mantle. This allowed us to treat cementless cups up to a size of 48 mm with the double socket technique. As a consequence, we lateralized the center of rotation of the hip joint by about 3 mm and left the edge of the DMC uncovered by the former cup, which may lead to edge loading. In the particular case of the early cup loosening, the undercoverage as well as the absence of shell roughening could have caused the failure.
Additionally, not only were several cups lateralized, but most of them were corrected in inclination and anteversion in order to achieve a more physiological position, especially in order to avoid posterior dislocation using the posterolateral approach. Whether this method is more prone to failure cannot be fully answered with the data available. Although 16 out of 17 DMCs showed no signs of loosening at FU, the sample size of the total collective is too small and the FU time too limited to allow for a final conclusion.
In comparison to the other studies (0–17%) we had a similar dislocation rate (7,7%). The two dislocations both occurred at the small DMC size of 43 mm. It is accepted that the smaller the DMC size, the smaller the jumping distance is until dislocation occurs [21]. This should be considered with regard when applying the double socket technique.
Possible discussed downsides apart from an expected increased polyethylene wear is the increased risk of taper corrosion. In a series of retrieved DMCs, Lombardo et al. found moderate fretting and corrosion damage around the trunnion as well as at the femoral head [23]. The changes were not correlated with the positioning of the implant according to the safe zones by Lewinnek et al. [24], although larger studies are needed to allow for a proper conclusion. Apart from galvanic corrosion which may occur with every modular head-taper interface, the use of a large head may lead to an increased friction at the taper. However, this could not be proven in a mathematic model with experimental validation [22]. As with conventional hip prostheses, an impingement between the taper and the cup can also lead to metal wear. In one case with a highly anteverted DMC, a beginning defect of the posterior aspect of the taper of a cementless titanium revision stem could be detected which was probably caused by posterior implant-to-implant impingement. The patient remained asymptomatic and had no dislocations at the latest FU and is closely monitored.
Our study has several limitations. First it is limited by its retrospective nature and its inherent limitations. Moreover, although to our knowledge being the largest study on this topic, it is still hampered by the low sample size. Although we followed up all patients alive, a considerable number had died at FU. A further limitation worth discussing is the omitted shell roughening of the primary implant as well as the use of the double socket technique for smaller primary implants or the use of larger DMCs with a reorientation which was not in line with the primary implant. Whether this variation is more failure-prone can only be evaluated with longer FUs with lager patient cohorts, possibly in a multi-center setting.