Fractures involving the posterior wall are the most common fractures of the acetabulum, accounting for approximately half of Letournel's originally published series.1,2 Despite their frequency, the outcome of patients afflicted by these injuries is often disappointing.3–6 The ability to predict poor outcome before initiating treatment would be useful for counseling patients and selecting appropriate treatment plans. Treatment options include nonoperative treatment, percutaneous fixation, open reduction internal fixation (ORIF), arthroplasty, and combinations of these techniques.2,7–11
There is a growing interest in performing acute total hip arthroplasty (THA) for acetabular fractures in cases in which ORIF is suspected to fail, particularly among elderly patients.9–11 This is particularly relevant, given the prevalence of geriatric acetabular fractures has more than doubled over the past 3 decades.12 Compounding the increasing incidence is the fact that elderly patients are more likely to have unfavorable fracture characteristics such as marginal impaction and comminution.12 A large meta-analysis demonstrated a 23% rate of conversion to THA among patients older than 55 years with acetabular fractures. Secondary conversion of an acetabular fracture managed with ORIF and later converted to a THA has been shown to have impaired outcomes and shorter survivorship compared with a primary THA for osteoarthritis.13,14
The purpose of this study was to identify risk factors for early conversion to THA in an effort to aid in counseling patients and selecting the optimal treatment. We had the following 3 hypothesizes: radiographic features of the injury would correlate with outcomes, fracture reduction quality would affect outcomes, and radiographic features of injury would correlate with reduction quality.
PATIENTS AND METHODS
After institutional review board approval, we searched our institutional trauma database for patients with acetabular fractures involving the posterior wall from 2005 to 2010 managed with ORIF. We included elementary and associated fracture patterns that had a posterior wall component managed using a Kocher–Langenbeck approach. Patients who did not have both preoperative and immediate postoperative computed tomography (CT) scans available for review and a minimum of 4 years of follow-up were excluded from the study. Eight patients did not have postoperative CT scans, of whom 6 were pediatric patients and 2 patients refused to be scanned. Of note, routine practice at our institution is to obtain postoperative CT scans of patients treated for pelvic ring and acetabular fractures.
A chart review was conducted to obtain baseline clinical data, and operative reports were reviewed. Mechanism of injury was categorized into 5 categories, “fall,” was considered a low-energy fall. Such as from standing or a step stool. The preoperative radiographs, preoperative CT scans, intraoperative imaging, and postoperative CT scans were reviewed by 2 orthopaedic trauma–trained surgeons. Preoperative imaging was reviewed for evidence of dislocation, comminution of the posterior wall (defined as more than 3 pieces), a femoral head lesion (includes impaction), impaction of the acetabulum, and presence of intraarticular loose bodies. Fragment of any size, loose body, or impaction/depression (loss of sphericity) that could be seen at the resolution of the CT scan on 2 consecutive slices was considered. Therefore, a 2-mm loose body was considered was considered equal to a 1-cm loose body. Given the small sample size, these characteristics were not subclassified. The postoperative CT scan was used to measure the quality of reduction. Specifically, the magnitude of the largest articular step-off and diastasis was recorded using the method described by Moed et al.4 If disagreement was noted between the 2 reviewers, then the primary author would review imaging and all 3 would come up with a consensus.
Given the broad catchment area of our institution, patients were contacted by telephone to inquire about secondary operative procedures and to administer the SF-8 and patient components of the Merle d'Aubigne Hip Scale (ROM component excluded). The primary outcome measure of the study was the rate of conversion to THA.
Data were analyzed using Stata version 13 (StataCorp, College Station, TX). Categorical variables were compared using Fisher exact test, whereas continuous variables were compared using Student t test. We performed a multivariate logistic regression to control for confounding among predictor variables. The variables found to be significant in bivariate analysis were added to the model using forward stepwise regression. We considered a significant difference if the P-value was less than 0.05.
Of the 150 patients with posterior wall acetabular fractures identified in our trauma database during the period of the study, 25 patients were excluded due to death, no postoperative CT scan, or age less than 18 years. This left 125 patients, of whom 65 patients (52%) were able to be evaluated in clinic or contacted by telephone more than 4 years postoperatively. The remaining 60 patients had follow-up that ranged from 6 weeks to 1.5 years (Fig. 1). The mean follow-up for the final study cohort was 6.9 years (range 4–9.3 years) after surgery (Table 1). The average age was 43 years and the majority of subjects were male (78.5%). There were 29 (44.6%) isolated posterior wall, 25 (38.5%) transverse posterior wall, 6 (9.2%) posterior column posterior wall, and 5 (7.7%) T-types with an associated posterior wall fracture. All patients underwent operative management within 2 weeks of injury.
The overall rate of conversion to THA was 16.9% (11/65). Five percent of patients required a THA at 2 years after surgery and 14% of patients required a THA at 6 years (Fig. 2). Based on a post hoc analysis, age greater than 46 years was associated with a 25.7% (9/35) rate of conversion to THA compared with 6.7% (2/30) for age 46 years or younger (P = 0.05). When considering annual incidence of conversion, patients older than 46 years of age were converted to THA at a rate 4.6-fold higher than younger patients (4.4% per year versus 1.0% per year, P = 0.036). Sex, fracture classification, time to closed reduction (13 patients had a dislocation), and time to surgery were not associated with the rate of conversion to THA (Table 1).
Radiographic features of injury including: the presence of dislocation, comminution of the posterior wall, femoral head impaction, acetabular impaction, and intraarticular loose bodies did not independently associate with the primary outcome measure. However, the presence of all 5 radiographic features concomitantly was associated with a 50% (5/10) rate of conversion to THA in contrast to 11% (6/55) if 4 or less features were present (P = 0.009). In addition, when group comparisons were made, statistical significance was noted between each group (P = 0.021) (Table 2). Furthermore, if patients had all 5 radiographic features, they had an annual rate of THA of 9.9% per year, and the medium group (2–4 features) had an annual rate of 1.7% per year (P = 0.01). No statistically significant difference was noted between the 0–1 features group compared with the medium group. Acetabular impaction was strongly associated with age greater than 46 years (89% versus 33%, P < 0.0001). Comminution, femoral head impaction, and other radiographic features were not associated with age. However, the combined radiographic score was strongly associated with age. There were no patients older than 46 years of age with 0/5 radiographic features compared with 13.3% in the younger cohort, whereas 25% of patients older than 46 years of age had 5/5 features compared with 3.3% in the younger cohort (P = 0.004). Despite more radiographically severe injuries among older patients, there was no difference in mechanism of injury between the 2 groups.
In terms of fracture reduction, among cases with less than 1 mm of diastasis and/or step-off on postoperative CT scan, there were no THA conversions (0/8) compared with 10% for 1–4 mm and 54% if either step-off or diastasis was 4 mm or more (P = 0.001) (see Figure, Supplemental Digital Content 1, http://links.lww.com/JOT/A509). The annual rate of THA for the reduction group that had a malreduction more than 4 mm was 10.6% per year, versus an annual rate of 1.4% for the reduction group that had a malreduction measuring between 2 and 4 mm (P = 0.0005). Furthermore, the presence of all 5 radiographic features of severe injury was associated with a reduction step-off or diastasis greater than 4 mm in 60% of cases (6/10) compared with 13% (7/55) of less severe injuries (P = 0.005). In multivariate analysis, only reduction quality was found to be a significant independent predictor of need for THA (odds ratio: 10.0; 95% confidence interval, 1.9–53.5; P = 0.007, Table 3). There was no difference in SF-8 (16.4 versus 17.4, P = 0.63) or modified Merle d'Aubigne scores (8.0 versus 8.9, P = 0.39) comparing patients who underwent THA and those who did not.
In multivariate analysis, only reduction quality was found to be a significant independent predictor of need for THA (odds ratio: 10.0; 95% confidence interval, 1.9–53.5; P = 0.007).
In a retrospective case series of patients managed operatively for acetabular fractures involving the posterior wall, we have demonstrated that high-energy mechanisms and a combination of radiographic features of severity portend a high rate of conversion to THA. Furthermore, we have shown that reduction step-off or diastasis greater than 1 mm, which cannot be detected with plain radiography, was strongly correlated with subsequent need for THA.
The findings of this study are not entirely unprecedented. In a series of 182 patients with posterior wall fractures treated operatively over a 20-year period in Toronto, Kreder et al15 showed a similar correlation between radiographic severity and subsequent outcome. Specifically, they demonstrated that posterior wall comminution, marginal impaction, and older age were associated with a higher rate of conversion to THA. However, they did not have postoperative CT scans for the majority of their patients and were not able to show a significant effect of postoperative reduction. In addition, their minimum follow-up was only 1 year with a mean of 4 years.
Moed et al4 included 67 patients in a study evaluating the effect of fracture reduction on the outcome of posterior wall acetabular fractures using the modified Merle d'Aubigne Hip Scale as a primary outcome at an average of 4 years after surgery. Although they did find that poor reduction was associated with lower outcome scores, the cutoff for reduction was a diastasis of more than 1 cm, whereas step-off was not found to be significantly associated with the outcome. These results strongly contrast with our study in which reductions of less than 1 mm had much greater longevity than fractures with step-off or diastasis of greater than 4 mm. We believe the longer duration of follow-up and more contemporary cohort with modern fine-cut CT scans may contribute to this discrepancy.
Authors from Greece reported the outcomes of 19 patients with posterior wall fractures with a minimum of 15 years of follow-up.16 The authors report only a single patient requiring conversion to THA and overall strong correlation between reduction quality and outcome. However, they did not use postoperative CT to evaluate reduction quality, and it is unlikely that the incidence and accessibility of THA for a patient population treated in the 1980s is equivalent to a more modern population.
There are several noteworthy limitations of this study. Foremost is our ability to only contact 52% of patients with a minimum of 4 years of follow-up, which results largely from the large geographic area served by our institution. We attempted to contact as many patients as possible by telephone, but were unsuccessful in reaching patients in many cases. We believe this is offset to some degree by the relatively long duration of follow-up (mean 7 years), use of a firm primary end point (conversion to THA), and validated patient-centered outcome instruments (SF-8 and Merle d'Aubigne hip score). Another important limitation is that the study was retrospective, which may limit data quality, particularly for many baseline clinical variables. However, because many of the baseline injury characteristics were obtained from digital radiographs, there is no bias due to recordkeeping error or poor recall. Because the majority of our patients did not have long-term radiographic follow-up, we could not comment on the reason for conversion to THA, such as avascular necrosis or posttraumatic arthritis. However, we were able to correlate the primary outcome with postoperative reduction quality but not with time to closed reduction, which could potentially suggest that failures occurred more often secondary to arthritis. Finally, we did not have adequate sample size to demonstrate the association between several variables with statistical significance nor were we able to conduct a multivariate analysis to identify the strongest independent risk factors for conversion to THA.
Posterior wall acetabular fractures associated with the combination of dislocation, comminution, intraarticular loose bodies, femoral head lesions, and acetabular impaction are associated with poorer reductions and higher rate of conversion to THA in comparison with less severe injuries. Patients should be counseled accordingly about the need for future arthroplasty, and surgeons can give consideration to primary THA in these severe cases. When ORIF is undertaken, anatomic restoration of the acetabulum to within 1 mm of both step-off and diastasis, which cannot be accurately detected with plain radiographs, is associated with the lowest risk of posttraumatic arthritis.
The authors thank Jessica Schisel for editorial assistance.
1. Judet R, Judet J, Letournel E. Fractures of the acetabulum: classification and surgical approaches. J Bone Joint Surg Am. 1964;46:1615–1646.
2. Letournel E, Judet R. Fractures of the Acetabulum. Berlin, Heidelberg: Springer Science & Business Media. 1993.
3. Moed B, McMichael J. Outcomes of posterior wall fractures of the acetabulum. J Bone Joint Surg Am. 2007;89:1170–1176.
4. Moed B, Carr S, Gruson K, et al. Computed tomographic assessment of fractures of the posterior wall of the acetabulum after operative treatment. J Bone Joint Surg Am. 2003;85:512–522.
5. Moed B, Carr S, Watson J. Open reduction and internal fixation of posterior wall fractures of the acetabulum. Clin Orthop Relat Res. 2000;57–67.
6. Moed B, WillsonCarr S, Watson J. Results of operative treatment of fractures of the posterior wall of the acetabulum. J Bone Joint Surg Am. 2002;84-A:752–758.
7. Matta J. Fractures of the acetabulum: accuracy of reduction and clinical results in patients managed operatively within three weeks after the injury. J Bone Joint Surg Am. 1996;78:1632–1645.
8. Enocson A, Blomfeldt R. Acetabular fractures in the elderly treated with a primary Burch-Schneider reinforcement ring, autologous bone graft, and a total hip arthroplasty: a prospective study with a 4-year follow-up. J Orthop Trauma. 2014;28:330–337.
9. Sierra R, Mabry T, Sems S, et al. Acetabular fractures: the role of total hip replacement. Bone Joint J. 2013;95-B:11–16.
10. Boraiah S, Ragsdale M, Achor T, et al. Open reduction internal fixation and primary total hip arthroplasty of selected acetabular fractures. J Orthop Trauma. 2009;23:243–248.
11. Herscovici D, Lindvall E, Bolhofner B, et al. The combined hip procedure: open reduction internal fixation combined with total hip arthroplasty for the management of acetabular fractures in the elderly. J Orthop Trauma. 2010;24:291–296.
12. Ferguson T, Patel R, Bhandari M, et al. Fractures of the acetabulum in patients aged 60 years and older: an epidemiological and radiological study. J Bone Joint Surg Br. 2010;92:250–257.
13. Schnaser E, Scarcella N, Vallier H. Acetabular fractures converted to total hip arthroplasties in the elderly: how does function compare to primary total hip arthroplasty? J Orthop Trauma. 2014;28:694–699.
14. Von Roth P, Abdel M, Harmsen W, et al. Total hip arthroplasty after operatively treated acetabular fracture: a concise follow-up, at a mean of twenty years, of a previous report. J Bone Joint Surg Am. 2015;97:288–291.
15. Kreder H, Rozen N, Borkhoff C, et al. Determinants of functional outcome after simple and complex acetabular fractures involving the posterior wall. J Bone Joint Surg Br. 2006;88:776–782.
16. Mitsionis G, Lykissas M, Motsis E, et al. Surgical management of posterior hip dislocations associated with posterior wall acetabular fracture: a study with a minimum follow-up of 15 years. J Orthop Trauma. 2012;26:460–465.