Standardized 3D-printed trabecular titanium augment and cup for acetabular bone defects in revision hip arthroplasty: a mid-term follow-up study

In accordance with the ethical standards of the committee on human experimentation of our institution, we retrospectively reviewed our hip revisions performed by a senior consultant orthopedic surgeon from 2015 to 2018. Patients were included if their clinical records and radiographs were consensually graded according to the Paprosky classification [13] of acetabular bone deficiencies as grade II or more. There were 59 total or partial (acetabular only) revisions in 59 patients were selected, and 1 was excluded due to incomplete clinical data. Seven different reconstruction techniques were applied for the remaining 58 patients (24 men and 34 women) (Table 1). The mean age at the time of revision was 61.7 (29–88) years. Of the 48 patients (48 hips) who were treated using standardized 3D‑printed TT cups and augments, 2 patients were lost to follow-up, and 3 died during the follow-up without undergoing further revision. Finally, 43 revisions in 43 patients were included. The patient flow diagram is shown in Fig. 1.

Demographic data for the 43 patients are described in Table 2. The indications for revision were aseptic loosening in 31 hips, progressive osteolysis in 5 hips, severe acetabular erosion after hemiarthroplasty in 3 hips, and 4 were the second stage of a two-stage revision for infection. All patients with aseptic failure reported symptoms of hip pain. The mean time elapsed from primary surgery to this revision was 9.4 (range 2.3–17.1) years.

The standardized 3D‑printed TT cups and augments (AK Medical, China) were manufactured by electron beam melting (EBM) technology and were composed of titanium particles melted layer by layer with a three-dimensional mesh structure. The pore size (600–800 μm) and porosity (80%) were determined with computer-aided design.

Computed tomography (CT) was performed to provide 3D structural information on acetabular bone deficiencies, and a 3D-printed pelvis model was established based on the CT data. Then, a simulated surgery was performed on the 3D model, and the preliminary decision of prosthesis selection, prosthesis sizing and the position of augmentation were made.

During the operation, the degree of the acetabular bone defect was confirmed and reassessed by comparison with the 3D-printed hip model. All 43 patients were operated on using a standard posterolateral approach. Extended trochanteric osteotomy (ETO) was performed in seven cases to remove the well-fixed femoral component because of malalignment of the stem (2 cases), leg length discrepancy (2 cases), impingement (1 case), and unacceptable abrasion or corrosion at the head-neck taper of the stem (2 cases). The other four patients underwent ETO for hip stiffness to assist exposure. After removal of the previous acetabular components, the lead surgeon debrided the remaining acetabulum. Synovial fluid and homogenate of the deep tissue specimens were obtained for all patients and were sent to the institutional laboratory for culture. For all patients with a positive culture, intravenous antimicrobial therapy was initiated to cover pathogens in accordance with the culture. The acetabulum defect was carefully and gently reamed from a smaller size to a diameter that was one millimeter or two millimeters less than the planned size to mesh with the augment tightly. The standardized 3D‑printed TT augment was inserted into the acetabular defect and was additionally fixed to the pelvis with 6.5-mm screws. Whenever possible, these screws were placed in the direction of the transmission of load, directly into the cranial ilium. After implantation and fixation of the augment, the acetabulum was reamed with the planned degrees of inclination and anteversion in successive increments from a smaller size to a diameter that was one millimeter or two millimeters less than the planned size. Post-reaming, the grinding debris was removed and repeatedly flushed. Morsellized allogeneic bone graft was used to fill the remaining defect. After that, a trial cup was tested and evaluated by intraoperative fluoroscopy. If the cup position, stability and contact area with the host bone were thought to be acceptable, the augment–cup interface was coated with polymethylmethacrylate cement. Then, a cementless standardized 3D‑printed TT cup was secured and impacted with an adequate press-fit and screws. Finally, the stability of the cup was evaluated again. If the femoral stem was loosening, the taper was significantly worn or osteolysis was present in the proximal femur, it was simultaneously revised. For the one case with pelvic discontinuity, a buttress was also used to construct the posterosuperior acetabular wall.

In general, patients were encouraged to walk with crutches, and partial weight-bearing was allowed on the first day after revision surgery. A full weight-bearing gait was permitted after clinical and radiological review at 6 weeks postoperatively. The patients were followed at regular intervals (6, 12, 24 weeks and annually thereafter). Standard anteroposterior (AP) and lateral radiographs were taken at three postoperative days and routinely at every visit.

Overall survival was defined as remaining free of any further revision of the acetabular component. The inclination and anteversion of the acetabular cup were measured on the AP radiograph based on the method of Bachhal et al. [14]. The Lewinnek safe zone [15] was used to assess the cup position. The criteria for osseointegration of uncemented acetabular components described by Moore et al. [16] were used to assess ingrowth. Loosening of the acetabular component was recognized by a change in its position. More than 3 mm migration of either horizontal or vertical orientation, radiolucent lines of 2 mm or more in all DeLee and Charnley zones [17], fracture of screws or variation in cup angle greater than 5° indicated radiographic failure. Radiographs were screened for progressive radiolucent lines (RLLs). The COR was measured on the preoperative and postoperative AP radiographs based on the modified method of Ranawat et al. [18]. The leg length discrepancy (LLD) was described as the discrepant distance from the base of the teardrop to the corresponding tip of the lesser trochanter.

Clinical evaluation was performed using the Harris hip score (HHS) and Oxford hip score (OHS). Health-related quality of life was assessed using the EuroQol five-dimensional questionnaire (EQ-5D).

All radiographs were reviewed by two of the authors (LJH and PFW) and were completed twice by each of the reviewers at least 3 weeks apart. Inter- and intra-observer reliability was assessed by intraclass correlation coefficients (ICCs) for the measurements. For all radiographic parameters, the mean correlation coefficient of the interobserver reliability was 0.88 (range, 0.83–0.94), and the mean correlation coefficient of the intraobserver reliability was 0.90 (range, 0.77–0.97).

The mean follow-up was 64.5 (range 49–84) months, with a minimum follow-up of 4 years. Acetabular revision was performed as a separate procedure in 13 (30.2%) cases (Fig. 2).

Of the 10 patients who underwent revision without 3D-printed cups and augments, one patient who underwent revision using a jumbo cup was lost to follow-up, and another patient who underwent IBG combined with a cemented cup died during the follow-up without undergoing further revision. One patient who was treated using an allograft combined with a TM cup and another patient who was treated using double TM cups failed due to aseptic loosening.