Abstract
Introduction
The aim of this study was to review the clinical, radiological and patient-reported outcomes with the use of cup-cage construct for pelvic discontinuity at our institution.
Methods
24 patients were identified at median 6-year (minimum 2 year, maximum 10 years) follow-up. 1 patient was converted to excision arthroplasty for infection. A further 3 patients required revision for instability but the cup-cage construct was not revised.
Results
We noted encouraging pain relief (mean WOMAC pain 85.6) and good functional outcome (mean WOMAC function 78.2, mean UCLA 5, mean OHS 78.6). Patient satisfaction with regards pain relief, function and return to recreational activities were noted to be good.
Conclusions
The cup-cage construct is a viable method of dealing with complex pelvic discontinuity. However, the failure rate due to loosening (4 cases) in this and other reports does prompt the need for further refinement of the technique and technology in this very challenging group of patients, as well as continued evaluation at the mid- and long-term so as to confirm the ongoing success of this method of reconstruction.
Introduction
Severe bone loss encountered during acetabular revision arthroplasty may be associated with dissociation between the proximal and distal halves of the acetabulum, referred to as pelvic discontinuity or dissociation.
Revision surgery for pelvic discontinuity in the presence of bone loss is challenging. Options include restoration of bone stock using morselised or structural bone graft, stabilisation of the fracture with 1 or more plates, or bridging of the defect with an ilioischial cage. Each of these are combined with revision of the socket. Results of these techniques have not been encouraging (1, 2). A more complex technique, when the bone loss is extreme is segmental replacement with massive segmental allograft or a massive endoprosthesis; so-called triflanged cups. The results following allograft implantation have not been encouraging (3), while those following the massive endoprosthesis have been more satisfactory (3).
The cup-cage reconstruction option has become popular for the management of pelvic discontinuity in recent years. This involves placing an enhanced-fixation porous revision tantalum shell, with or without bone graft or augments, in contact with the available bone stock and fixed with screws. A bridging cage placed within the tantalum shell, spanning the area of bone loss and the discontinuity, and fixed to the ilium and ischium, functions by decompressing the interface between the new fibre-metal shell and bone, improving the likelihood that the shell will osseointegrate. As the shell gains sufficient fixation to bone, the cage is offloaded and this prevents failure of the construct as previously seen with conventional cage alone reconstructions. The cup-cage construct provides robust fixation into which a cemented acetabular cup is placed in the required orientation. Results of this technique have been encouraging (4-5-6). While mid-term survival and hip outcome scores have been reported for this technique, there is insufficient literature on patient-reported outcome measures and satisfaction using this technique.
The aim of our study is to review our results of cup cage reconstruction for nontumour related pelvic discontinuity associated with bone loss. We reviewed clinical, radiological, and patient-reported outcomes at a minimum 2-year and median 6-year follow-up.
Methods
From 2004 to 2012, 24 cup-cage reconstructions were performed at our institution for arthroplasty-related pelvic discontinuity. This data was identified from our institutional database and our clinical ethics research board approved the study. Any cases related to a bone tumour were excluded from this review. Minimum 2-year clinical and radiological data was available in all patients. The median follow-up was 6 years (range 2-10 years, mean 5 years, standard deviation [SD] 3).
5 patients died of unrelated causes between 2 to 6 years after surgery (mean age at surgery 81 years [range 76 to 86 years]) and the latest quality of life measures were not available in these patients. A further 2 patients were lost to follow-up after their 2-year clinical and radiological appointments and were also not available for the most recent quality of life measures. In the 7 patients (5 deaths, 2 loss to follow-up) not available for final follow-up, the demographic data; surgical data; latest clinical data available were included in this study.
Demographic data of the study cohort are summarised in Table I. The mean age was 72 years (SD 13.6). The median body mass index (BMI) was 26.7. The Amerian Society of Anesthesiologists (ASA) Class was 1 in 1 patient; 2 in 13; 3 in 9; and 4 in 1. §3 patients were undergoing their first acetabular revision; 8 patients their second; and the remaining 3 were undergoing their third; fourth and fifth acetabular revisions respectively.
Demographic data
BMI = body mass index.
Of the revised acetabular cups, 7 were cemented and the rest were uncemented.
A concomitant femur revision was undertaken in 5 hips. In 2, a cemented CPT (Zimmer) stem was used. An S-ROM stem (Depuy Synthes), a Revitan stem (Zimmer) and a GMRS proximal femoral replacement stem (Stryker) were used in the 3 other hips.
Pelvic discontinuity was diagnosed on preoperative radiographs in all cases and confirmed intraoperatively by examination for movement between the proximal and distal parts of the acetabulum after removal of the acetabular shell.
Surgical technique
The acetabulum is adequately exposed, typically using a posterior approach, and the floor and rim are cleared of any debris and fibrous tissue. Tissue clamps placed on the proximal and distal acetabulum can be used to demonstrate movement of the 2 parts to confirm discontinuity. The remaining bone stock is lightly reamed and a Trabecular Metal Revision (TMR) shell (Zimmer) is inserted to find adequate posterior and superior support. We also displace or modestly distract the fracture fragments enough to provide ligamentotaxis or soft tissue stability and enhanced fixation. This manoeuvre increases the initial bone implant interface stability. Prior to shell insertion, morselised allograft mixed with any autograft from reamings may be packed into contained defects by gentle reverse reaming to produce new acetabular bone stock, and Trabecular Metal augments (Zimmer) may be used for uncontained defects, in addition to the Trabecular Metal Revision cup. An adequate area of host bone devoid of autograft or allograft is retained in contact with the cup to allow for eventual bone ingrowth and permanent fixation
The TMR shell (Zimmer) is placed in an attitude of approximately 50° lateral opening and neutral to slight retroversion and then fixed with multiple screws; preferably diverging into the available bone stock. This attitude of the shell is critical to allow appropriate placement of the cage and intimate contact of the superior flange with the outer table of the ilium. A “trial run” with the shell in place and the trial cage aligned over top serves to choose the correct shell alignment, after which screw fixation of the shell is completed.
Next, a cage (Zimmer) of suitable proximal blade length is chosen to decompress the underlying construct during weight-bearing. The inferior keel of the cage is placed into a slot within the ischium; the cavity of their cage fitting into the cavity of the TMR shell; and the superior blade conforming with the outer table of the pelvis. The cages are specific for each TMR shell depending on the shell size. When the ischium is adequate, the inferior keel of the cage is slotted into it, else it may have to be placed onto the remaining ischium, after care is taken to locate and mobilise the sciatic nerve. Rigid fixation of the cage is achieved using screws into the ilium. Screws may also be placed between the cage and the TMR shell for additional fixation. It is not necessary to align the screw holes in the cup with the screw holes in the cage. A new screw hole in the cup can be made through a screw hole in the cage using a 4-mm burr. All screw heads are covered with bone wax before preparation to cement the cemented cup into the cage, in case screw removal is needed at a later date.
Following this, a matching polyethylene cemented cup of suitable diameter, which is specifically designed for use with these cages, is fixed in place using antibiotic-loaded cement. The final attitude of the cemented cup within the cage must be optimised to achieve the desired orientation (40° of lateral opening and 20°-25° anteversion) and to correct for the attitude of the TMR cup, which might not have been placed in an ideal position in order to accommodate the cage.
A trial reduction is then done to confirm leg length, horizontal offset, stability and range of movement.
Postoperative mobilisation is dependent on the stability of the construct achieved on the table, but as a general rule, patients are mobilised toe-touch weight-bearing bearing for 6 weeks followed by partial weight-bearing for 6 more weeks and check radiographs are evaluated at that point before progression to full weight-bearing.
Patient-reported outcomes measures and radiological review
We reviewed clinical, radiological and patient-reported outcome measures (PROM's) of our study cohort at minimum 2-year follow-up. Radiological evaluation consisted of antero-posterior (AP) and iliac-oblique views of the pelvis and a lateral view of the hip. We used Gill's criteria for radiological evaluation of the cages (2). Definite loosening is suspected if the screws are broken or if there is acetabular migration >5 mm, or if a complete, progressive radiolucent line was present medial and superior to the cage, or around the screws. Probable loosening is suspected if there is progressive radiolucency medial or superior to the cage. Definite or probable cage loosening was considered a radiological failure. Progressive migration of the cup was also considered a radiological failure.
Clinical failure was defined as revision of the acetabular component due to any cause. PROM's assessment at the time of the study was performed using Western Ontario and McMaster Universities Arthritis Index (WOMAC), SF12, Oxford Hip Score, Harris Hip Score, satisfaction scale and University of California Los Angeles (UCLA) scores in all cases. The satisfaction scale was individually measured for pain, function, recreation and overall satisfaction on a scale of 1 to 4 (very satisfied, somewhat satisfied, somewhat dissatisfied and very dissatisfied.)
Statistical analysis was performed using SPSS version 15.0 (SPSS). Data was expressed as means and standard deviations.
Results
Bone graft was used in 13 hips (morselised femoral head allograft in 11 cases and packed cancellous freeze-dried bone graft in 2 hips). The median diameter of the TMR shell used was 64 mm (range 56-76 mm). The long iliac flange cage was used in 4 hips and the short flange in the rest. A 36-mm diameter head was used in 6 hips and, a 28-mm diameter head was used in 1 hip, and a 32-mm diameter head was used in the remaining 17 cases. Figure 1 shows a successful cup-cage reconstruction at mid-term follow-up.
Plain radiograph demonstrating a successful cup-cage reconstruction for pelvic discontinuity at mid-term follow-up.
The cup-cage construct was revised in 1 patient for infection. He developed Escherichia coli infection 1 month after surgery and underwent excision arthroplasty. He elected not to undergo further reconstruction and died of a myocardial infarction 6 years after surgery.
3 patients had revision procedures for recurrent instability. In 2 patients the cemented cup was changed to a constrained cemented cup 1 year after surgery and in 1 patient (2 years post surgery) the proximal modular component of the femur (GMRS [Stryker Orthopaedics]) was lengthened. Revision of the cup-cage construct was not necessary in any of these cases.
Adverse radiological changes were noted in further 4 patients. A 9-mm superior migration of the cup-cage construct was noted in 1 patient at 1-year follow-up but remained stable and was not revised. In 1 patient, fracture of the ischial spike was noted 1 year after surgery but no further change has been noted on follow-up radiographs. Breakage of the cage screws were noted in 2 patients at 1 year and 3 years post surgery but no migration of the construct was noted in any of the cases. The remaining cases showed no signs of radiological failure of the acetabular construct.
The PROMs at minimum 2 years (median 6 years, maximum 10 years) were available in 18 patients (9 patients excluded – 5 deaths, 2 loss to follow-up and 2 revision of cup-cage constructs). The results are summarised in Table II.
Quality of life and functional outcome measures
WOMAC = Western Ontario and McMaster Universities Arthritis Index; UCLA = University of California Los Angeles activity score.
We noted encouraging overall pain relief (mean WOMAC pain 85.6) and good overall functional outcome (mean WOMAC function 78.2, mean UCLA 5); and quality of life measures (mean SF-12 PC 40.8; mean SF-12 MC 60.4). 77.8% of the patients had WOMAC pain and global scores above 75 and 66.7% of the patients had WOMAC function score above 75.
The mean Oxford score in the 4 patients who showed adverse radiological changes of construct migration or hardware failure of was noted to be lower at 57 (range 48-81.3, SD 21.2) and the mean WOMAC function score was 59.9 (range 50-75, SD 13.3). They had acceptable pain [WOMAC pain 60 (range 50-80, SD 17.3)] and satisfaction scores [WOMAC satisfaction 79.2 (range 75 −87.5, SD 7.2)]. For this reason, they need to be considered as failures with unlikely long-term durable results.
If one considers aseptic failure as any revision plus clinical/radiological failures the success rate of this procedure is 75% (18/24). Half of the failures were instability and the other 3 were patients with poor clinical scores plus radiological loosening. These 3 patients elected not to have revision to date.
Discussion
Pelvic dissociation or discontinuity associated with failed acetabular components is a complex management problem. In our series of 24 cases that were treated with a cup-cage construct for pelvic discontinuity, we have demonstrated encouraging satisfaction and pain relief and good functional outcome at median 6 years (minimum 2 years) of follow-up. 1 revision of the construct was performed for infection.
3 cases in our series required revision of the cemented acetabular cup for instability; although the cup-cage construct was not revised. A constrained acetabular liner was used successfully in 2 patients and subtotal revision of the modular femoral component was undertaken in 1. We do not routinely use a constrained acetabular shell unless irremediable instability of the construct is identified on the table. This reluctance is based on the increased forces relayed to the bone implant interface with a constrained construct and the risk of the entire construct loosening or displacing. In addition, four patients showed signs of radiographic failure although they elected not to have revision surgery.
Dealing with pelvic discontinuity continues to be a challenge. When dealing with discontinuity in the presence of bone loss, use of isolated bone grafting has high failure (7). Traditional cages used alone have shown high failure rate (8). Recent studies have shown promising results with the use of the cup-cage construct (5, 9). The distraction technique (5, 10) is another option with good early outcomes. The technique proposes the addition of even more distraction or fragment separation than we routinely use with the cup-cage technique in our hands. Custom-made triflange constructs also currently available (11-12-13) and early reports have shown promise.
Kosashvili et al (6) reported minimum 2-year (2-6 years) follow-up of 26 cup-cage acetabular reconstructions performed in patients with pelvic discontinuity and massive bone loss. They noticed that 88.5% (23 hips) demonstrated no evidence of clinical or radiographic loosening. Three acetabular components demonstrated migration of >5 mm and 2 required revision surgery. The Harris Hip Scores improved from 47 points preoperatively to 77 points postoperatively. Complications included 2 dislocations and 1 infection. Abolghasemian et al (4) reported the outcome of 26 pelvic discontinuities in 24 patients) treated by the cup-cage technique at a mean follow-up of 82 months (12-113 months) and compared them with a series of 19 pelvic discontinuities in 19 patients treated with a cage at a mean follow-up of 69 months (1-170 months). They documented 4 failures of the cup-cage group (15%) and 13 (68%) of the cage group due to septic or aseptic loosening. The 7-year survivorship was 87.2% for the cup-cage group and 49.9% for the cage-alone group (p = 0.009). Radiological union of the discontinuity was found in all successful cases in the cup-cage group and 3 of the successful cage cases. 3 hips in the cup-cage group developed early radiological migration of the components, which stabilised with a successful outcome. Sporer et al (10) reported an alternative management of chronic pelvic discontinuity with the use of acetabular distraction using porous tantalum elliptical acetabular component (with or without augments) in 28 patients. At minimum 2-year follow-up (average 4.5 years; range 2-7 years) 1 patient required re-revision for aseptic loosening. 15 patients remained radiographically stable while 4 patients had early migration with subsequent radiological stabilisation. The modified Merled'Aubigne'-Postel score at latest follow-up was 6.6 (range 3.3-9.6).
The mid- to long-term results of the cup-cage reconstruction for acetabular reconstructions with massive bone loss remains to be established. There is also paucity of literature around PROMs and patient satisfaction, after cup-cage reconstruction. Most published literature has used Harris Hip Score or its modifications (3). With increasing number of revision operations being undertaken in the setting of deficient bone stock, we believe it is important to document patient-reported outcomes and satisfaction with revision surgery. Our study has demonstrated 83.3% mean patient satisfaction and mean WOMAC pain score of 85.6. The mean WOMAC functional score was 78.2 and mean UCLA scale was 5. This is in keeping with the age group of the study cohort and complexity of the reconstruction. The high satisfaction reflects the ability of these patients to carry on with day-to-day activities without pain. It was our observation that none of the patients were engaging in impact related recreational activities; however, we had a high proportion of patients participating in swimming. 77.8% of the patients in our cohort had WOMAC pain and global scores above 75 and 66.7% of the patients had WOMAC function score above 75. This indicates that the majority of patients in this study reported only mild pain and mild functional disability. Further follow-up of the patients with lower WOMAC scores is essential to document their outcome.
In conclusion, the cup-cage construct is a promising method of dealing with complex pelvic discontinuity. However, the failure rate due to loosening (4 cases) in this and other reports does prompt the need for further refinement of the technique and technology in this very challenging group of patients, as well as continued evaluation at the mid- and long-term so as to confirm the ongoing success of this method of reconstruction.
Footnotes
Financial support: None.
Conflict of interest: None.