1 Introduction

Acetabular reconstruction in cases of significant bone loss and pelvic discontinuity poses a major challenge in revision surgery. Treatment strategies includes the use of large hemispherical cups with augments,¹ acetabular impaction bone grafting and cemented cups,² oblong components,³ reconstruction rings or cages⁴ and acetabular distraction technique.⁵ The clinical outcomes of these complex cases have been variable with high failure rates due to lack of host bone to ensure adequate stability and fixation of the components.^2,6

More recently the use of custom triflange acetabular components have gained popularity.^7,8 The main indication for these implants is the presence of massive periacetabular bone loss and/or pelvic discontinuity (Paprosky 3A/3B and AAOS IV) that precludes the ability to obtain stable fixation with hemispherical cups.^9,10 A custom triflange components allow bridging of bony defects to obtain fixation to the remaining host bone; ilium, ischium and pubic bones. Component flanges are designed with optimal geometry and orientation to provide an intimate fit against the host bone and bridge existing bone defects to facilitate initial fixation. This allows better restoration of hip biomechanics and reduces the risk of component migration and early failure.^7,8,11,12 Custom triflange components have shown excellent survivorship at short and medium term follow up, although longer term follow-up is needed to determine the effects of these rigid constructs on host bone.^12–14 Further, they are expensive implants that require complex preoperative planning, manufacturing process, iatrogenic bone loss can lead to imperfect fitting and high rates of instability have been reported in some series.^12,15

The aim of this study was to evaluate our short to medium term outcomes of revision hip replacements for significant bone loss and pelvic discontinuity at our unit using custom triflange acetabular components.

2 Materials and methods

Our unit is a specialist tertiary hip reconstruction centre with high volume revision hip surgery with an average 300 hip revision surgeries performed per year (between 1st April 2015–31st March 2018; 908 hip revisions were undertaken- National Joint Registry data¹⁶). Our indications to use custom triflange components were cases with significant periacetabular bone loss (Paprosky 3A/3B) and pelvic discontinuity (AAOS IV) in which conventional reconstruction techniques were felt to have a high likelihood of failure by the operating surgeons.

A retrospective consecutive review of all patients undergoing custom triflange acetabular revision with a minimum 24 months follow up were included. All patients had detailed preoperative assessments including, when indicated, blood markers and synovial fluid aspirations to rule out periprosthetic joint infection. Detailed radiographic examinations were obtained including computed tomography (CT). Bony defects were classified according to Paprosky and AAOS classification systems and confirmed intraoperatively.^9,10

All patients underwent a thin-cut (1 mm) CT scan of the pelvis which was sent to the manufacturers where a three-dimensional reconstruction of the images was created (Fig. 1d). From this, a computer-generated model of the bone stock was created and a prosthesis designed. Two designs of prosthesis were used. The Biomet Patient-Matched Implants triflange acetabular components (Biomet Inc, Warsaw, IN, USA) were used in the earlier cases. These are milled from solid titanium alloy and subsequently porous plasma spray and hydroxyapatite coated. In later cases, the Materialise Patient-Specific Hip Implants (Materialise NV, Leuven, Belgium) prosthesis were used. These implants are manufactured using additive manufacturing (AM) technology with selective laser melting which uses a focused laser beam to melt titanium alloy (Ti6Al4V) powder layer-by-layer. The implant is designed to have a highly porous titanium inner surface in contact with the host bone to facilitate on-growth. The inner patient-specific instruments manufactured using medical grade epoxy monomer were produced using the same AM technology. This offers 3D-printed drill guides which allows placing the required cross-fixated screws as indicated by the detailed preoperative plan.

Close

Fig-1 (a): Preoperative anteroposterior radiograph of the pelvis (patient 5) with bilateral bipolar hemiarthroplasties and left IIIB acetabular defect; (b) plain radiograph with stable construct at 6 months follow up; (c) at 16 months follow up there is a broken ischial screw but overall stable construct, the patient is mobilising with one stick and under close monitoring. (d): example of 3-D (CT-based) reconstruction of the same patient used in preoperative planning and manufacturing process of the custom implants with prototyping technology. (e): example of quality of bone for screw fixation (CT-based) which aids planning of screw placements. (f): example of preoperative plan provided by the manufacturer with proposed plan for screw placement with appropriate length to ensure adequate purchase based on bony quality.

Export to PPT

Surgeries were performed by the senior authors (NS, HWJ) through an extended posterior approach. The acetabulum was exposed, and dissection extended into the ilium wing; anteriorly to the pubic bone and posteriorly exposing the ischium. Failed components removed and bony surfaces prepared as per preoperative planning for triflange cup positioning and screw placements. Sclerotic bony surfaces were curetted and minimally reamed minimising host bone loss. Bone grafting, allograft, also added to fill in any small voids between the implant and host bone particularly medially to reconstruct medial defects and to bridge pelvic discontinuity defects.

Postoperatively, patients underwent routine revision arthroplasty rehabilitation protocol including prophylactic antibiotics and thromboembolic prophylaxis, were mobilised weight bearing as tolerated. Clinical and radiographic assessments were obtained at 6 weeks, 3, 6, 12 months and annually thereafter. The primary outcome measure was the rate of radiographic failure. This was defined by the presence of component migration or progressive radiolucent lines around the component-bone interface.¹⁷ Complications were also recorded.

3 Results

Seventeen patients (17 hips) underwent revision surgery between 2013 and 2017 and were all included in this study. There were 3 males/14 females with a mean age of 72 years (range 61–83). The average follow-up was 3.6 years (2-7 years). All patients had multiple previous surgeries (Table 1). Bony defects were classified as Paprosky 3B in 13/17 hips (76%) with pelvic discontinuity encountered in the majority of cases 15/17 hips (88%) and intra-pelvic failed components in 11/17 (64%) (Figs. 2–4) (Table 2). At final follow up, no radiographic failures were observed although 3 patients developed complications (17.6%); one patient had post-operative haematoma requiring washout out; one patient had intra-operative ilium fracture however the implant was stable and no implant migration or failure was noted but remains under close monitoring. Finally, one patient had recurrent dislocation and underwent revision with a change of modular liner to a constrained liner and higher offset stem. Radiographic analysis of all patients showed stable constructs in all patients, however, 3 patients had once screw breakage with no radiolucent lines around the implants or migration. Patients are clinically asymptomatic and (Table 3) and remain under close monitoring.

Close

Fig-2 (a) Preoperative anteroposterior radiograph of the pelvis (patient 8) with significant acetabular defect (Paprosky IIIA) of right hip with catastrophic failure, metal debris from previous surgeries and proximal migration; (b) 30 months follow up radiograph following reconstruction with a stable construct, areas of the iliac defect were bone grafted but has now partially resorbed.

Export to PPT

Close

Fig-3 (a) Preoperative anteroposterior radiograph of the pelvis (patient 15) with right pelvic discontinuity intrapelvic socket; (b/c) Plain and oblique radiographs at 4 year follow up with a stable construct.

Export to PPT

Close

Fig-4 (a) Preoperative anteroposterior radiograph of the pelvis (patient 17) with pelvic discontinuity of the right hip and intrapelvic socket. (b) Plain radiograph at 5 year follow up with a stable construct.

Export to PPT

4 Discussion

Traditional acetabular reconstruction techniques with large hemispherical cups or augments to achieve stability on host bone have had excellent long-term results for small defects.¹⁸ Similarly, impaction bone grafting and cemented polyethylene components have also had satisfactory long-term outcomes.¹⁹ Acetabular distraction technique popularised by Dr Paprosky^5,21 where porous tantalum components are used to achieve both initial stability and subsequent long-term biological fixation with good clinical results reported at medium-term follow up. However, failures have been observed when these techniques were applied in cases that extended their indications.²⁰

Therefore, in cases where defects are too big to allow stable fixation of the implants, these defects are bridged by a custom implant with flanges spanning the ilium, ischium and pubic bone providing triangular fixation points to the remaining host bone.

Further, these flanges allow biological ongrowth on host bone for long term stability. Other advantages of custom triflange components are improved conformity with host bone and greater construct rigidity and resistance to fatigue failure.²² On the other hand, the main disadvantages of custom implants are the requirement for advanced imaging, manufacturing time, expense and the inability to modify the implant intraoperatively.²³

Our indications for using these custom implants are pelvic discontinuity and in cases where other reconstruction techniques are likely to fail such as some Paprosky 3A and 3B defects. In this study, all but two patients had pelvic discontinuity and over 60% having intra-pelvic failed components reflecting the complexity of these cases. The few case-series published in the literature are summarised in Table 4. Most of these studies included patients with varying severities of bone loss but reported overall similar outcomes but with overall high rates of complications and component failures. To date, we have not revised any triflange cups for failure. Although we did revise a modular liner into a constrained liner in one patient for dislocation. Further, three patients have had broken screws, singular screw, although currently asymptotic. One patient has a loose iliac screw, one with a broken pubic and one with ischial screw. All patients remain under close monitoring for any further screw breakage which might indicate pending failure, although the radiographic features for the patient with a broken ischia screw have been static over the last 5 years.

This study is limited by the small number of patients included and the relatively short follow up, however this is the first UK based series on this challenging group of patients using custom triflange acetabular components. Our early experience in using custom made triflange acetabular component for patients with significant bone loss and pelvic discontinuity is promising but longer-term follow-up is needed to assess its long-term survivorship and clinical outcomes.