Determining Stability in Posterior Wall Acetabular Fractures

ORIGINAL ARTICLE

Determining Stability in Posterior Wall Acetabular Fractures Reza Firoozabadi, MD, MA,* Clay Spitler, MD,† Calvin Schlepp, MD,‡ Benjamin Hamilton, MS,§ Julie Agel, MA, ATC,k Milton “Chip” Routt, MD,¶ and Paul Tornetta, MD**

Objectives: To determine if the radiographic parameters of femoral head coverage by the intact posterior wall, acetabular version, and location of the fracture or a history of dislocation were determinates of hip stability in patients with posterior wall acetabular fractures.

Design: Retrospective review.

Key Words: posterior wall, acetabular fractures, instability, acetabular dome, version, dislocation, examination under anesthesia, trauma

Level of Evidence: Diagnostic Level II. See Instructions for Authors for a complete description of levels of evidence. (J Orthop Trauma 2015;29:465–469)

Setting: Level I trauma hospital. Patients: One hundred eighty-five consecutive patients with isolated unilateral posterior wall (OTA 62-A1) acetabular fractures. Intervention: Patients underwent dynamic stress fluoroscopic examination under general anesthesia to determine hip stability. Main Outcome Measurements: A number of radiographic measurements were performed, and an examination under anesthesia served as a standard to compare stable versus unstable hips. Results: Examination under anesthesia (EUA) determined 116 hips to be stable and 22 hips as unstable. Moed and Keith method of wall size measurements and cranial exit point of fracture was statistically different between stable and unstable hips. Twentythree percent of the unstable hips had wall sizes less than 20%. Average cranial exit point of fracture from dome was 5.0 mm in the unstable group and 9.5 mm in the stable group, and fractures that extend into the dome demonstrate a statistically significant increase in hip instability.

Conclusions: Determination of hip stability can be challenging in patients with posterior wall acetabular fractures. Our data suggest that the location of the exit point of the fracture in relation to the dome of the acetabulum is a radiographic marker that can be used to aid physician in determining stability, and wall sizes less than 20% is not a reliable indicator of stability. Accepted for publication April 22, 2015. From the *Department of Orthopaedic Surgery, Harborview Medical Center, Seattle, WA; †University of Mississippi, Jackson, MS; ‡University of Washington, Seattle, WA; §Case Western School of Medicine, Cleveland, OH; kHarborview Medical Center, University of Washington, Seattle, WA; ¶University of Texas, Houston, TX; and **Boston University Medical Center, Boston, MA. Presented in part at the Annual Meeting of the Orthopaedic Trauma Association, October 17, 2014, Tampa, FL. The authors report no conflict of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jorthotrauma. com). Reprints: Reza Firoozabadi, MD, MA, Department of Orthopaedic Surgery, Harborview Medical Center, Box 359798, 325 Ninth Avenue, Seattle, WA 98104-2499 (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Orthop Trauma Volume 29, Number 10, October 2015

INTRODUCTION Fractures of the posterior wall of the acetabulum are the most common injuries to the hip socket, comprising 20%–30% of all acetabular fractures.1 Nonoperative management of these injuries is supported when the hip is stable and congruent.2,3 This occurs when the native anatomy and remaining wall size are large enough to maintain the femoral head within the acetabulum.4 Traditionally, the posterior wall fragment size was used as a predictor of hip stability.5–7 Broadly speaking, hips with a “large” wall fragment measuring greater than 66% as measured by the technique of Calkins et al, 40% as described by Keith et al, and exceeding 50% as described by Moed et al are deemed to be unstable and require open reduction internal fixation (ORIF).5–7 Fractures affecting less than 20% of the posterior wall as described by Keith et al or Moed et al have a high likelihood of being stable and theoretically can be treated nonoperatively. Fracture sizes in between the “small” and “large” group have been labeled as indeterminate. Davis and Moed8 determined that experts in the field of acetabular surgery could not predict stability for this indeterminate group using plain radiographs and computed tomography (CT). Furthermore, recent evidence has shown that “small” posterior wall fragments less than 20% of the wall can cause the hip to be unstable.3,9 Therefore, radiographic measurements of wall size have only been marginally successful in accurately predicting instability (Fig. 1). Although wall size has been studied extensively and is a critical component of acetabular fracture stability, many other anatomic factors need to be considered given the complex bony anatomy of the hip joint. Based on our clinical experience, we wanted to determine if other fracture characteristics such as cranial exit point of the fracture, acetabular version, lateral center edge angle, percent of femoral head coverage, and history of dislocation would assist in predicting hip stability in patients with posterior wall acetabular fractures.

PATIENTS AND METHODS Institutional review board approval was obtained, and a retrospective review identified 504 consecutive posterior www.jorthotrauma.com |

465

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Firoozabadi et al

FIGURE 1. Two different patients with posterior wall acetabular fractures. The patient on the left has a stable left hip with a relatively large posterior wall acetabular fracture, and the patient on the right had an unstable hip with a relatively small posterior wall acetabular fracture. Editor’s note: A color image accompanies the online version of this article.

wall fractures of the acetabulum (OTA 62-A1) from our prospectively gathered database of patients collected at a single level I trauma center from December 2001 to July 2013.10 Chart review was performed to identify injuries that underwent dynamic examination of the affected hip in the operating theater as a test of hip stability. Patients younger than 18 years of age and patients with bilateral acetabular injuries were excluded. One hundred thirty-eight cases were identified, 22 of which has positive dynamic stress tests. After identification of these 138 cases, charts and radiographic studies were retrospectively reviewed to characterize the patient demographics, mechanisms, and the radiographic characteristics of the acetabulum and posterior wall fracture. All patients had plain radiographs or volume-rendered anteroposterior reconstructions that were obtained before dynamic hip examination. Preference was taken to use plain film imaging to measure the lateral center edge angle as described by Wiberg.11 The angle is formed by the intersection of a line drawn through the center of the femoral head and extending to the lateral edge of the sourcil and a line perpendicular to one joining the 2 femoral head centers. Acetabular version was assessed on CT using the roof edge (RE) angle and the equatorial edge (EE) angle as


described by Reynolds et al and applied to acetabular fractures by Werner et al.12,13 In brief, these angles are measured between a reference line, drawn between the center of the sacrum and the center of the pubic symphysis representing the axial center of the pelvis, and a line drawn between the anterior and posterior acetabular rims either at the level of maximal femoral head diameter (EE angle) or the opening at the most cranial extent (RE angle). Radiographic descriptions of posterior wall fracture size using two-dimensional (2D) CT images have been previously described in multiple studies (Fig. 2). We measured the size of the fracture using 4 methods: Calkins, Keith, Moed, and our groups “percent femoral head coverage.” The Calkins method described the “approximate acetabular fracture index.”5 This ratio of intact posterior wall to uninjured posterior wall is determined by drawing a straight line between the medial and lateral margin of the posterior articular surface at that level of largest fracture involvement irrespective of the level. This is compared with the same line on the injured side at the same acetabular level. The range deemed indeterminate to predict hip instability was an intact wall of 34.3%–55.2%, the width of the uninjured wall. The Keith method measures the medial to lateral fragment width at the level of the fovea.6 The comparison measurement is medial to lateral distance from the medial wall to posterior rim at the level of the fovea of the uninjured acetabulum. The ratio is then reported representing the fragment size as percentage of intact acetabulum. Indeterminate range for instability reported by Keith et al is between 20% and 40% in their cadaver-based study. The Moed method is measured and reported similarly to the Keith method, but measurements are made at the level of greatest fracture disruption of the posterior wall.7 This method was developed to provide a better representation of fracture fragments that involve more acetabulum either above or below the fovea. The indeterminate fracture fragment size range is reported as 20%–50%. A fourth method was developed to describe the size of the posterior wall fracture’s effect on acetabular depth and posterior coverage of the femoral head (also known as “percent femoral head coverage”). This measurement is taken at

FIGURE 2. Three methods used to calculate posterior wall fracture fragment size. Far left, Calkins et al method, which measures the smallest amount of intact acetabular arc. A straight-line medial–lateral measurement is made of the remaining intact articular posterior wall acetabular segment at the level of the greatest amount of the fracture involvement (X). The length of the posterior acetabular arc is determined from the contralateral uninjured hip at the same level (Y). X divided by Y multiplied by 100 provides the index. Middle panel, Keith et al method, which measures wall size at the level of the fovea. The depth of the fracture fragment is measured at the level of the fovea (X). The percentage of the fragment size is calculated from the ratio of the measured depth of the fracture side to intact matched contralateral acetabular depth (Y). Far right, Moed et al method, which is a modification of Keith method, which measures at the level of the largest wall deficit instead of measuring at the fovea. Reproduced with permission from Moed BR et al.7 Copyright @ 2009, Wolters Kluwer Health, Inc.

466

| www.jorthotrauma.com

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.



the level of the largest femoral head spherical diameter, typically at or near the fovea. A reference line is drawn between the anterior acetabular rim and the intact remnant of the posterior wall. Perpendicular to the reference line, the maximum distance of femoral head medial to the line and the diameter of the femoral head provide the ratio of femoral head coverage in the plane of the intact acetabulum (Fig. 3). In all measurements, marginal impaction was considered a component of the fracture fragment. The cranial/caudal position of the fracture was determined by locating the cranial most exit point of the fracture line as it exits out of the articular surface of the acetabulum. It is important to distinguish that the cranial exit point does not refer to the extra-articular fracture line, as the extra-articular component can extend well above and/or well below the joint involvement. The dome of the acetabulum was designated as the reference point with the fracture exiting through the dome as 0 mm. The exit point was identified using axial images and cross-referenced with sagittal and coronal slices on the supine CT scan after the hip was reduced if it was dislocated. A dynamic stress examination under anesthesia was confirmed in review of the operative reports by a traumatrained orthopaedic surgeon. For the examination, the patients were in a supine position with the hip in the neutral rotational position. The hip was then gradually flexed, followed by internal rotation, and finally progressive manual force in line with the femur in the flexed position while anteroposterior and obturator oblique images were obtained. Any fluoroscopic evidence of hip subluxation and joint incongruence was considered a positive test for instability. Spot fluoroscopic images were initially obtained followed by live fluoroscopy. All unstable fractures were treated operatively unless the patient had a medical condition that did not permit surgery. In the majority of unstable fracture cases, the patients underwent ORIF at the time of the EUA. Some patients underwent

FIGURE 3. Calculation of percent femoral head coverage. This measurement is taken at the level of the largest femoral head diameter. A reference line is drawn between the anterior acetabular rim and the intact remnant of the posterior wall. Perpendicular to the reference line, the maximum distance of femoral head medial to the line and the diameter of the femoral head provide the ratio of femoral head coverage in the plane of the intact acetabulum. Percent femoral head coverage = [X/(X + Y)] · 100. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Posterior Wall Acetabular Fracture Stability

delayed ORIF for the following reasons: EUA performed during the weekend or at night or EUA performed with another procedure that was more urgent and patient not deemed medically stable for further operative management. Rehabilitation protocol for nonoperative stable patients and operative unstable patients was the same. Patients are allowed weight of limb weight-bearing and isometric exercises for the first 6 weeks with hip flexion precautions. At 6 weeks, radiographs were obtained and progressive weight-bearing as tolerated with limb/buttock/abdominal/lumbar core strengthening exercises guided by a physical therapist was prescribed. Subsequently, patients returned to clinic at 12 weeks postoperatively for follow-up radiographs and returned to work soon thereafter. Student t test was used to assess for statistical significance of continuous variables, including fracture site exit point, center edge angle, acetabular version, and fracture size. Chi-square testing was used to assess categorical measurements of wall stability when assessing small, indeterminate, and large wall sizes and the presence of dislocation in relation to stability. Significance was set to P , 0.05 for all tests.

RESULTS Of the 138 patients who sustained unilateral posterior wall acetabular fractures and had an EUA, 116 (84%) were stable and 22 (16%) were unstable. Motor vehicle collision was the most common injury mechanism, and the rate between the stable and unstable groups was not statistically significant (see Table, Supplemental Digital Content 1, http://links.lww.com/BOT/A398). Sixty-nine of the 116 patients (59%) in the stable group had a dislocation that was documented, and 15 of the 22 patients (68%) had a dislocation in the unstable group (P = 0.49). Displaced wall size as described by Moed, Keith, and Calkins is reported in Table 1. Although the techniques described by Moed and Keith for measuring displaced wall size demonstrated statistical significance between the groups, Calkins’ did not, and the 2 groups had nearly identical wall sizes on average using Calkins’ technique. The smallest displaced wall size noted to be unstable was 15% using the techniques by Moed and Keith. The largest displaced wall size that was stable was 65% using the above techniques. Cranial exit point of the fracture was measured on CT as the distance between the axial slice containing the acetabular dome and the most cranial axial slice of the acetabulum that demonstrated a fracture line. Cranial exit point of the fracture in the unstable group ranged from 18.75 to 0 mm from the acetabular dome and from 48.75 to 0 mm in the stable group. The mean exit point of the fracture in the unstable group was 5.0 mm from the acetabular dome, which

TABLE 1. Wall Size Data

Moed Keith Calkins

Stable, %

Unstable, %

Stable Range, %

Unstable Range, %

P

21 17 25

26 27 25

2–64 0–56 2–72

15–48 16–26 9–48

0.02 0.001 0.96

www.jorthotrauma.com |

467



Firoozabadi et al

was significantly more cranial than the mean exit point of the stable group at 9.5 mm (P = 0.004). Further analysis revealed that all fractures that exited within 5 mm of the dome were significantly more likely to display instability at EUA (P = 0.037). However, when including all fractures within 10 mm of the dome, there was not a significant increase in the incidence of instability. Furthermore, fractures that extended into the dome demonstrated a statistically significant increase in hip instability (P = 0.006). RE and EE angles were used to assess acetabular version on axial CT images. RE angle average measured 5.28 in the unstable group and 4.88 in the stable group and was not statistically significant (Table 2). EE angle was also not statistically significant between unstable and stable groups as depicted in Table 2. Average lateral center edge angles were nearly identical in both groups, at 40.38 in the unstable group and 40.48 in the stable group. A subgroup analysis was performed on wall sizes that measured less than 20% by the Moed technique. Five of the 22 patients (23%) in the unstable group had wall sizes less than 20%, with a range of 15%–19% for wall size (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A399). Four of the 5 had a cranial exit point of the posterior wall fracture within 5 mm of the dome. Of the 21 patients who had a stable hip on EUA and a wall size less than 20%, 90% had a cranial exit point more than 5 mm from the acetabular dome. Interestingly, 4 of the 5 patients with wall sizes less than 20% had a history of dislocation.

DISCUSSION Hip stability after a posterior wall acetabular fracture is difficult to predict, and the gold standard remains examination under anesthesia. Hip instability after a closed reduction of a posterior wall acetabular fracture is an indication for surgery.14,15 CT using 2D axial imaging has been the mainstay of determining the location and size of the posterior wall acetabular fracture. Postreduction imaging that displays an incongruent hip joint, intra-articular debris, and/or a fracture involving more than 50% of the posterior wall has generally been regarded as an unstable joint, which requires surgery.5,7,14–16 Unfortunately, a large number of acetabular fractures do not fulfill these criteria, and the stability of the joint is extremely difficult to determine. Traditionally, fractures involving less than 20% of the posterior wall have been deemed as stable. Our clinical experience and recent evidence have shown this not to be the case.9 Although using the cutoff of 20% may seem reasonable with a prediction rate of 92%, the other 8% of patients will most likely have a poor clinical outcome with an unstable hip.8 Furthermore, the study by Davis and Moed supports the notion that experts in the field of acetabular fracture care have a difficult

time determining stability for fractures that involve 20%–50% of the posterior wall using imaging modalities. Many surgeons use a history of dislocation to assist them in determining the stability of the hip joint in cases where wall size does not provide them a concrete answer. Although a history of dislocation would seem intuitive, it is not a reliable marker of instability.3,8,17 As a result, we as orthopaedic surgeons are poor predictors of hip stability in posterior wall acetabular fractures that comprise less than 50% of the posterior wall. Tornetta suggested that all patients without clear indication for operative intervention have a dynamic stress examination under fluoroscopy to confirm hip stability, and in those that were stable, the results were as good as those of operative treatment. The purpose of our study was to find additional radiographic markers that could be used to assist in determining hip instability in patients with posterior wall acetabular fractures. If any specific measurements correlated well with either stability or instability, then those patients could avoid stress examination and be treated with ORIF or observation with confidence. Our hypothesis that acetabular version would correlate with instability was not supported by our study. Although varying acetabular version has been associated with different types of acetabular fractures, we measured version at 2 distinct axial points (RE angle and EE angle at the fovea) and neither point was statistically significant in term in relation to hip instability.13 A potential confounding variable in measuring acetabular version using these techniques is that we could not correct for pelvic tilt. The study by Dandachli et al showed that correction of pelvic tilt correlated 2D images better with 3D images when measuring version. We attempted to correct for pelvic tilt but were not able to do so due to fact that the data required to make the correction on the CT scans were purged on a routine basis.18,19 Despite this, hip instability on examination would likely be treated operatively irrespective of pelvic tilt. Additionally, in the coronal plane, center edge angle was also not statistically significant. To the best of our knowledge, this article is the first to assess posterior wall stability using the cranial extent of the fracture in relation to the acetabular dome. We are by no means the first to study fracture relationship in regard to vicinity to the acetabular dome. Rowe and Lowell first reported that fractures of the acetabulum that did not involve the weight-bearing area had better results.20 This was followed by authors defining the minimum area of the unaffected acetabular roof necessary to obtain acceptable results without surgery. They used axial CT scans to measure the subchondral arc in relation to the weight-bearing zone.17,21 The weight-bearing surface of the acetabulum was presumed to be intact if measurement of the roof arcs was greater than or equal to 45 and the subchondral CT arc of 10 mm was not broken. Although roof and subchondral arcs can be used for

TABLE 2. Acetabular Version and Center Edge Angle RE angle Equatorial angle Center edge angle

468

Stable, degrees

Unstable, degrees

Stable Range, degrees

Unstable Range, degrees

P

4.8 12 40

5.2 13 40

218 to 32 25 to 25 25 to 67

210 to 21 210 to 24 21 to 58

0.85 0.69 0.97

| www.jorthotrauma.com




a majority of acetabular fractures, they do not apply to fractures of the posterior wall, which occur in a region beyond the plane of measurement.3 However our study suggests that the location of the cranial exit point of the posterior wall acetabular fracture is significant. Fractures deemed unstable based on EUA were found to have a cranial exit point closer to the dome, compared with stable fractures. Specifically, unstable injuries had on average a cranial exit point within 5.0 mm of the dome and stable injuries had an average cranial exit point of 9.5 mm from the dome. Interestingly, the subgroup analysis of fractures that would have been predicted to be stable based on a wall size of less than 20%, 23% (5 patients) were found to be unstable. Four of the 5 patients (80%) had a cranial exit point within 5 mm of the acetabular dome. Our findings indicate that potentially small posterior wall fragments can be unstable, and these tend to occur in patients with cranial fracture exit points within 5 mm of the acetabular dome. Using the criteria of cranial exit point, surgeons could potentially avoid misclassifying an unstable hip joint as stable. Regarding the 2D CT assessment of posterior wall fracture fragment size, our finding found significant differences between the 3 techniques in the ability to differentiate between unstable and stable hip joints. The method of Calkins et al did not produce statistically significant differences between the unstable and stable groups. This method uses the level of the largest fracture fragment and the length of the posterior acetabular arc on the intact contralateral hip. Like other authors, we found the measurement of the contralateral hip very challenging and questionable in some cases because we were not sure we were measuring at the exact same location of the acetabulum on both sides because of the patient not being perpendicular to the scanner gantry.6–8 The method by Keith et al and Moed et al did produce statistically significant different results and are easier means of measuring the fragment size. We prefer the Moed technique due to the fact that it measures the fragment size at the largest location of the fracture and not just at the fovea. Finally, we did not find a history of hip dislocation as an indicator of hip instability. Cranial exit point of posterior wall acetabular fracture can potentially be used to help determine stability of posterior wall acetabular fractures. Specifically, fractures that enter the dome are at high risk for instability. We advocate the use of dynamic stress fluoroscopic examination of the hip in patients who have a cranial exit point within 5 mm of the acetabular dome, even if the fracture fragment is less than 20% of the posterior wall. We advocate using the term “indeterminate” to fractures that are less than 50% of the posterior wall and not just fractures that involve


Posterior Wall Acetabular Fracture Stability

20%–50% of the posterior wall. This study does not support the use of hip version or center edge angle in determining hip stability. REFERENCES 1. Letournel E, Judet R. Fractures of the Acetabulum. New York, NY: Springer; 1993. 2. Moed BR, Spoonamore MJ. Management of Acetabular Fractures. Oxford Textbook of Orthopedics and Trauma. Oxford, UK: Oxford University Press; 2002:2182–2201. 3. Tornetta P III. Non-operative management of acetabular fractures. The use of dynamic stress views. J Bone Joint Surg Br. 1999;81:67–70. 4. Grimshaw CS, Moed BR. Outcomes of posterior wall fractures of the acetabulum treated nonoperatively after diagnostic screening with dynamic stress examination under anesthesia. J Bone Joint Surg Am. 2010;92:2792–2800. 5. Calkins MS, Zych G, Latta L, et al. Computed tomography evaluation of stability in posterior fracture dislocation of the hip. Clin Orthop Relat Res. 1988;227:152–163. 6. Keith JE Jr, Brashear HR Jr, Guilford WB. Stability of posterior fracturedislocations of the hip. Quantitative assessment using computed tomography. J Bone Joint Surg Am. 1988;70:711–714. 7. Moed BR, Ajibade DA, Israel H. Computed tomography as a predictor of hip stability status in posterior wall fractures of the acetabulum. J Orthop Trauma. 2009;23:7–15. 8. Davis AT, Moed BR. Can experts in acetabular fracture care determine hip stability after posterior wall fractures using plain radiographs and computed tomography? J Orthop Trauma. 2013;27:587–591. 9. Reagan JM, Moed BR. Can computed tomography predict hip stability in posterior wall acetabular fractures? Clin Orthop Relat Res. 2011;469: 2035–2041. 10. Marsh JL, Slongo TF, Agel J, et al. Fracture and dislocation classification compendium—2007: Orthopaedic Trauma Association classification, database and outcomes committee. J Orthop Trauma. 2007;21:S1–S133. 11. Wiberg G. Studies on dysplastic acetabula and congenital subluxation of the hip joint. Acta Chir Scandinavica. 1939;83:58S. 12. Reynolds D, Lucas J, Klaue K. Retroversion of the acetabulum. A cause of hip pain. J Bone Joint Surg Br. 1999;81:281–288. 13. Werner CM, Copeland CE, Ruckstuhl T, et al. Acetabular fracture types vary with different acetabular version. Int Ortho. 2012;36:2559–2563. 14. Baumgaertner MR. Fractures of the posterior wall of the acetabulum. J Am Acad Orthop Surg. 1999;7:54–65. 15. Larson CB. Fracture dislocations of the hip. Clin Orthop Relat Res. 1973: 147–154. 16. Vailas JC, Hurwitz S, Wiesel SW. Posterior acetabular fracturedislocations: fragment size, joint capsule, and stability. J Trauma. 1989;29:1494–1496. 17. Matta JM, Anderson LM, Epstein HC, et al. Fractures of the acetabulum. A retrospective analysis. Clin Orthop Relat Res. 1986:230–240. 18. Dandachli W, Ul Islam S, Tippett R, et al. Analysis of acetabular version in the native hip: comparison between 2D axial CT and 3D CT measurements. Skeletal Radiol. 2011;40:877–883. 19. McArthur B, Cross M, Geatrakas C, et al. Measuring acetabular component version after THA: CT or plain radiograph? Clin Orthop Relat Res. 2012;470:2810–2818. 20. Rowe CR, Lowell JD. Prognosis of fractures of the acetabulum. J Bone Joint Surg Am. 1961;43:30–59. 21. Olson SA, Matta JM. The computerized tomography subchondral arc: a new method of assessing acetabular articular continuity after fracture (a preliminary report). J Orthop Trauma. 1993;7:402–413.

www.jorthotrauma.com |

469
