Introduction

Total hip arthroplasty (THA) is among the most successful procedures in orthopedic surgery in the treatment of symptomatic osteoarthritis regarding the improvement of the quality of life, pain relief and restoration of function [1]. Hip replacement has consistently increased over the past decade with more than 1 million arthroplasties currently performed every year [2]. Despite the excellent outcomes and low complication rate following THA, the number of revision THA cases increases directly proportional to the increasing number of primary THA [3,4,5]. However, revision THA is associated with longer hospital length of stay, increased operative time, increased blood loss, higher complication rates and higher mortality [3, 6, 7].

Using an internervous, intermuscular plane, the direct anterior approach (DAA) provides a true minimally invasive interval to the hip for primary THA [8,9,10,11,12]. In primary elective THA, the DAA has shown excellent functional and radiographic outcomes, as well as a low dislocation rate [10, 13, 14]. Its soft-tissue-preserving nature results in a significant reduction in postoperative pain and recovery time [15]. However, nowadays, the most frequently used approaches for primary THA and revision surgery are the more extended posterior and lateral approaches; whereas, the DAA is often considered for primary THA only. This concern is, among others, based on the assumption that the distal extension of the DAA is limited due to associated damage of neurovascular structures [16]. Thus, exposure might be insufficient for revision surgery [16, 17]. Additionally, the limited visualization of the acetabular cup and the femoral stem as well as the technical difficulty to remove the femoral stem through this approach further discourages surgeons from using the DAA in the revision setting. Up to date, there are only limited data available concerning revision hip arthroplasty through the DAA [18, 19]. Therefore, the purpose of the present study was to report the peri- and postoperative complications as well as the clinical and radiographic outcomes of revision THA performed through the DAA, and compare our data to the existing literature.

Materials and methods

This study was approved by the institutional review board and the ethical committee. Each patient provided written informed consent before participation. It was conducted entirely at the authors' institution. Adult patients who underwent revision THA through the DAA and had completed at least 1-year follow-up at the time of data collection were included. All procedures were performed by two orthopedic consultants, experienced in hip revision surgery as well as in THA through the DAA. Procedures were categorized as follows: Liner and head exchanges, acetabular cup revisions, femoral stem revisions, and revision of both components during one procedure.

Patient characteristics

The medical records of all patients who underwent THA at our institution from January 2009 to August 2017 were retrospectively examined. From this database, a total of 117 revision THA have been extracted. 54 patients (46%) underwent a revision THA through the extended lateral approach, mostly because this approach was used for implantation of the primary prosthesis. In other cases, the lateral approach was selected because of a long stem prosthesis in situ, which was not amenable for extraction through the DAA. 63 patients (54%) underwent revision THA through the DAA: 21 liner and head exchanges, 26 acetabular cup revisions, 13 femoral stem revisions, and 3 revisions of both components within one procedure. 27 male patients and 36 female patients have been included with an average age of 73 years (range: 35–93 years). The average clinical follow-up was 18 months (range 12–72 months). Baseline characteristics including demographic information, comorbidities likely to influence postoperative complications and the patients' physical status according to the American Society of Anaesthesiologists (ASA) were recorded [20] (Table 1). Data were further reviewed for perioperative information including cause of revision, procedure, operative time and estimated blood loss.

Table 1 Patient characteristics, comorbidities and postoperative functional score
Full size table

Clinical evaluation

Patients were routinely followed up clinically and radiographically after 3 months, 1 year and 2 years following the revision surgery. Further follow-ups depended on the overall follow-up period.Medical records including operative reports, hospital records, and outpatient clinic notes were reviewed for readmission, complication, re-operation and Harris Hip Score (HHS) at every visit. An independent orthopedic surgeon at our clinic performed each examination in a standardized matter. For the current study, significant complications were considered: superficial and deep infection according to the criteria of the musculoskeletal infection society [21]; dislocation and periprosthetic fractures; polyethylene wear; aseptic loosening; hematoma and iliopsoas impingement that required additional intervention.

Radiological measurements

Two separate orthopedic residents individually evaluated the postoperative X-rays of all patients, who underwent acetabular cup revision. Each radiograph was obtained in a standardized manner with the lower limbs held together in a neutral position and the anterior superior iliac spine parallel to each other on the X-ray table. Pre- and postoperative radiographs of each patient were compared to evaluate changes in the horizontal and vertical center of rotation (COR), which was defined as the distance of the COR preoperative to the distance of the COR postoperative [22]. A negative value indicated a medialized or cranialized COR. Furthermore, the acetabular inclination was measured in the anteroposterior view as the angle between the acetabular axis and the horizontal axis and acetabular anteversion in the lateral view as the angle between the acetabular axis and the coronal axis [23].

Surgical technique

The technique of revision THA through the DAA builds on the same surgical approach used in primary THA. Patients are placed in the supine position using a traction table (AMIS® Mobile Leg Positioner, Medacta International SA, Castel San Pietro, Switzerland), which allows performing revision THA with a single, scrubbed assistant. The patient is prepared from the ipsilateral knee to the inferior costal margin, which allows both proximal and distal extension of the approach as needed. After performing a V-shaped capsulotomy, transection of the reflected head of the rectus femoris muscle allows for better exposure of the superior lateral surface of the ileum and the anterior aspect of the acetabular rim. Before removal of the femoral head, an extended capsulotomy is required to mobilize and externally rotate the femur sufficiently.

Acetabular revision

In isolated acetabular cup revision or exchange of modular parts solely, the exposure is practically no different from that during primary THA. However, some critical steps might improve the exposure of the acetabulum. An extended capsular release is important to prevent periprosthetic fracture during the mobilization of the femur. The well-fixed femoral stem is then retracted laterally, and the acetabulum is assessed over the medial aspect of the femoral component. To improve the acetabular exposure, Hohmann retractors can be inserted over the anterior and inferomedial acetabular rim. Additionally, exposure can be enhanced by slightly flexing the hip. The acetabular cup can be removed with a curved chisel or a special acetabular cup removal system. To get access to the lateral aspect of the ileum, the tensor fascia latae muscle (TFL) can be released from its insertion on the ilium and retracted lateral.

Femoral stem revision

At the authors institution, the femoral diaphysis is accessed at the posterior border of the vastus lateralis muscle, since splitting the vastus lateralis muscle or using the interval between the rectus femoris muscle and the vastus lateralis muscle endangers innervating branches of the femoral nerve to the structures lateral to the incision, thereby affecting the vastus lateralis and the lateral portions of the vastus intermedius [16, 17, 24]. If revision of the femoral component is performed, it is often necessary to have a direct approach down the femoral medullary canal to safely exchange the femoral stem. The exposure can be improved by placement of the patient using a leg positioner on both legs. The knee of the affected leg is flexed to 90°, and the hip is maximally externally rotated, lowered and adducted across the other side of the body below the opposite knee. Before extraction of the femoral stem, it is crucial to completely expose the medial and lateral side of the implant to avoid fractures of the femoral shaft and the greater trochanter. If the stem still cannot be safely removed, a V-shaped, biplanar (Chevron Style) Osteotomy of the great trochanter can be performed [25, 26]. The osteotomy should be performed as long as possible to increase the healing surface area. It is important to preserve an intact muscle-osseous sleeve, connecting the vastus lateralis muscle, the greater trochanter, and the gluteus medius muscle. After removal of the implant, fixation of the greater trochanter is not mandatory after implantation of another femoral stem, since the osteotomy will often reduce anatomically after restoration of the hip length and offset, due to its intact muscular sleeve (Fig. 1) [25].

Fig. 1
figure 1

a Preoperative anteroposterior radiograph of the hip with aseptic loosening of the left femoral component. b Anteroposterior radiograph after revision of the femoral component of the left hip through the DAA with a trochanteric osteotomy. After reconstruction of the hip length and offset, the osteotomy reduced anatomically without further fixation

Full size image

Results

Perioperative parameters

The average operative time and intraoperative blood loss varied depending on the procedure performed. Surgical time ranged from 80 min (range 34–108 min), if the liner and head were exchanged, and 130 min (range 91–195 min), if the acetabular cup and the femoral stem were revised. The average intraoperative blood loss was 436 ml (range 100–2000 ml). Table 2 provides further information on perioperative parameters in the different subgroups.

Table 2 Perioperative parameters
Full size table

Indication for revision

The two most common diagnoses for modular liner and head exchanges were infection (76%) and dislocation of the hip (24%). In case of hip dislocation, a longer head with equal diameter was implanted to lengthen the leg, thereby increasing soft tissue tensioning to achieve a more stable prosthesis. Revision of the femoral component was performed because of Vancouver-type B2 periprosthetic femur fractures in 46% (6 cases) and aseptic loosening in 31%. Treatment consisted of removal of the stem, reduction and tension band wiring of the fracture and insertion of a longer stem. In three patients, an extended trochanteric osteotomy was necessary to remove the femoral component. In two patients, the trochanter was left without fixation, and in one case, additional fixation with a tension band construct was performed. In nine cases (69%) of revision of the femoral or both components, the femoral stem was augmented with cement to enable early mobilization with full weight-bearing due to advanced patient age. Table 3 provides a summary of the principal diagnosis, which led to the revision in the different subgroups.

Table 3 Reason for revision total hip arthroplasty
Full size table

Functional outcome

The mean HHS at one year postoperative was 91 (range 74–100). In all subgroups, the HHS indicated good to excellent results at one year postoperative. Table 1 lists the mean HHS in the different subgroups.

Complication and re-operation rate

At a mean follow-up of 18 months, the overall complication and re-operation rates were 14.3% and 12.7%, respectively. Further information regarding complication and re-operation rates are listed in Table 4.

Table 4 Summary of complications
Full size table

Radiographic findings

26 patients underwent acetabular cup revision. The position of the acetabular component was confirmed with intra-operative fluoroscopy at the time of surgery. The average acetabular inclination was 42° (range 30°–61°) with 86.2% of the acetabular components in the Lewinnek “safe zone” for inclination. The mean acetabular anteversion was 21° (range 5°–34°) with 68.9% of the components in the Lewinnek “safe zone” for anteversion. The mean medialization and cranialization of the COR were 4 mm (range 2–11 mm) and 5 mm (1–9 mm), respectively.

Discussion

Due to its use of a true inter-muscular and inter-nervous plane and the preservation of the external rotators as well as the hip abductors, the DAA shows worldwide growing popularity in primary THA [10, 27,28,29]. Nevertheless, the available data in the literature regarding revision arthroplasty using the DAA are in a rudimentary stand. Therefore, it is important to further improve the knowledge of the DAA in the revision setting. The purpose of the study was to examine the perioperative and postoperative complications as well as clinical and radiological outcomes of revision THA performed through the DAA.

In the presented series, acetabular cup revision through the DAA was possible with adequate component positioning regarding inclination and anteversion as well as reconstruction of the COR. The functional outcome in this subgroup was highly satisfactory with an average HHS of 89 (range 74–100). Similarly, in a retrospective study, Horsthemke et al. reported an almost anatomically reconstruction of the COR after acetabular cup revision through the DAA with 87.5% of the components lying within the Lewinnek “safe zone” of inclination [30]. Together, these data suggest that the DAA allows for adequate positioning of the acetabular component even in the revision setting with good functional outcomes.

Femoral revision through the DAA is more complicated as the maneuvers for removal of the femoral stem carry an increased risk of iatrogenic fracture of the greater trochanter or even the femoral shaft [31, 32]. Our data showed a complication rate of 15.3%. However, due to the limited number of stem revisions in our cohort, adequate statistical interpretations should be viewed with caution. Additionally, it should be taken into account that most of the revised femoral components in this study were short stem designs, specifically designed for primary implantation through the DAA. Furthermore, in 77% of the cases, the fixation of the stem into the femur was already debilitated due to periprosthetic fractures or aseptic loosening. Therefore, our findings might not be applicable to other stem designs such as straight, rectangular stems and revision of the femoral component through the DAA should only be considered in selected cases. Furthermore, due to advanced patient age in nine cases, cemented stems have been implanted, allowing for early mobilization with full weight-bearing. Although uncemented implants for femoral revision THA are more popular, recent data show excellent long-term results regarding survivorship of cemented femoral components used for revision THA [33].

In our cohort, revision THA because of periprosthetic infection was performed in 17 cases. 16 patients were treated with exchange of the mobile parts, whereof 6 patients needed recurrent soft-tissue revision to eradicate the infection. Despite the repeated surgery, patients showed excellent functional outcomes with a mean HHS of 93 (range 86–100). Therefore, we consider the DAA as a valuable option for the treatment of acute periprosthetic infection.

The overall complication and re-operation rates of this study are comparable to the current literature on revision THA (Fig. 2). In a retrospective analysis of 1366 patients undergoing revision THA using the direct lateral approach, Jafari et al. reported a total complication rate of 18.6% [34] and Badarudeen et al. described in a retrospective analysis of 3555 revision THA through various approaches an overall re-operation rate of 15.8% at 1-year follow-up [35]. As in our study, the most frequently reported complications after revision THA are periprosthetic infection and dislocation [34, 35]. In prior studies on Medicare population, the rate of dislocation ranged from 5.3 to 20% after revision THA [35,36,37]. In a retrospective analysis of revision THA through the posterior approach, Dimitriou et al. reported a dislocation rate of 9% [38]. However, other authors showed a decrease of the dislocation rate down to 1.9% after revision THA through the posterior and posterolateral approach, if the soft tissue was repaired during the procedure [39, 40]. In our series, two anterior hip dislocations (9.5%) occurred after exchange of the liner and femoral head, which accounts for an overall dislocation rate of 3.2% in the whole cohort. The reason for the primary revision in these cases was earlier hip dislocation. In both cases, the acetabular component showed increased anteversion. Retrospectively, an acetabular cup revision with conversion to a dual-mobility cup system might have been the appropriate treatment in this setting. Consistent with our findings, Cogan et al. and Horsthemke et al. reported a dislocation rate of 6.6% and 0% after acetabular cup revision through the DAA [30, 41]. Together, these data indicate similar dislocation rates after revision THA through the DAA compared to revision surgery through the posterior approach with posterior capsular repair.

Fig. 2
figure 2

Summary of complications after revision THA through the DAA

Full size image

As described in the literature, revision THA involves a high risk of infection. Badarudeen et al. reported an infection rate of 17.3% after revision THA [35] and infection was determined as the most common cause of failure after revision THA through the direct lateral approach [34]. Using the posterior approach for revision THA, Liow et al. and Dimitriou et al. reported a total infection rate of 2% [38, 42]. In this study, deep infection occurred in three patients (4.7%), which is comparable to previous data of Kennon et al., who reported an infection rate of 2.5% after revision THA through the DAA [18].

This study should be interpreted in light of its several limitations. A significant limitation is the small number of patients in the different subgroups and the heterogeneity of the presented data, as the causes of revision THA in the different subgroups were miscellaneous. Another limitation is the retrospective study design. However, due to the standardized clinical and radiological follow-up protocol and the excellent documentation through the orthopedic surgeons of our institution, most of the patient data we needed were available for the current analysis.

Apparently, there is a significant skepticism to use the DAA for revision surgery. Our data suggest that the DAA might be a credible option for acetabular revision or exchange of the mobile parts. Revision of the femoral component through the DAA depends on the stem design and is feasible in selected cases. However, surgeons should be aware of the technical difficulties to remove the femoral stem and be prepared to extend the approach distally or even perform a trochanteric osteotomy if required.