- Email: [email protected]
The effect of early operative stabilization on late displacement of zone I and II sacral fractures
Injury, Int. J. Care Injured 44 (2013) 199–202
Contents lists available at SciVerse ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
The effect of early operative stabilization on late displacement of zone I and II sacral fractures Osa Emohare a,b,c,*, Nathaniel Slinkard b,c, Paul Lafferty b,c, Sandy Vang b,c, Robert Morgan a,b,c a
Center for Spine and Spinal Cord Injury, Regions Hospital, St Paul, MN, United States Department of Orthopedic Surgery, University of Minnesota, Minneapolis, MN, United States c Department of Orthopedics, Regions Hospital, St Paul, MN, United States b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 4 November 2012
Introduction: This study was designed to evaluate the effect on displacement of early operative stabilization on unstable fractures when compared to stable fractures of the sacrum. Methods: Patient consisted of those sustaining traumatic pelvic fractures that also included sacral fractures of Denis type I and type II classification, who were over 18 at the time of the study. Patients were managed emergently, as judged appropriate at the time and then subsequently divided into two cohorts, comprising those who were either treated operatively or non-operatively. The operative group comprised those treated with either internal fixation or external fixation. Results: Twenty-eight patients had zone II fractures, and 20 had zone I fractures. Zone II fractures showed average displacements of 6.5 mm and 6.9 mm in the rostral–caudal and anteroposterior directions, respectively, at final follow up. Zone I fractures had average displacements of 6.6 mm and 6.1 mm in both directions. There were no significant differences between zone I and II sacral fractures (rostral–caudal P = 0.74, anteroposterior P = 0.24). Average changes in fracture displacement in patients with zone I fractures were 0.6–1.0 mm in both directions. Average changes in zone II fractures were 1.8– 1.5 mm in both directions. There were no significant differences between the average changes in zone I and II fractures in any direction (rostral–caudal P = 0.64, anteroposterior P = 0.68) or in average displacements at final follow up in any of zone or the entire cohort. Statistically significant differences were noted in average changes in displacement in zone II fractures in the anteroposterior plane (P = 0.03) and the overall cohort in the anteroposterior plane (P = 0.02). Conclusion: Operative fixation for unstable sacral fractures ensures displacement at follow up is comparable with stable fractures treated non operatively. ß 2012 Elsevier Ltd. All rights reserved.
Keywords: Fracture fixation Fractures Malunion Orthopaedics Pelvis Sacrum Sacral fractures Denis classification Operative management of sacral fractures Non operative management of sacral fractures
Introduction The optimal approach for the management of sacral fractures has evolved from being predominantly non operative to current practice, where operative intervention is often preferred for displaced fractures.1 Originally reported by Malgaigne in 1876, sacral fractures comprise up to 45% of all pelvic fractures and are commonly caused by high-energy injuries. In a study of 776 pelvic fractures, 236 sacral fractures were found to be present in association with pelvic fractures.2 With an associated pelvic fracture, sacral fractures can often be missed, with up to 50% having been missed historically.3 Although advances in imaging are now improving this rate, significant proportions are still identified late.3
* Corresponding author at: Department of Orthopedics, Regions Hospital, St Paul, MN, United States. Tel.: +1 651 447 1941. E-mail address: [email protected] (O. Emohare). 0020–1383/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.injury.2012.11.004
Three distinct patterns of sacral fractures have been observed and these form the basis for classifying fractures according to the Denis system of classification, as each group has unique characteristics.2 The most commonly observed fracture type is the zone I fracture, which occurs lateral to the sacral foramen; these fractures are more often stable and generally, but not exclusively, treated non operatively. The second most common fracture type is classed as the zone II fracture, which extends through the sacral foramen. Zone II fractures are less stable than zone I fractures and consequently are associated with an increased rate of mal-unions. The third type is the zone III fracture, which is found medial to the sacral foramen and characterized by the presence of concomitant neurological injury in most cases. Studies have shown that patients with fracture displacements >0.5–1 cm have been associated with a higher incidence of non-union, malunion, and long term post operative back pain.4,5 Sacral fractures that initially appear minimally displaced have recently been shown to be at significant risk for late displacement. Bruce et al. found that incomplete lateral compression sacral
200
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
fractures associated with ipsilateral rami fractures can be treated nonoperatively and are unlikely to displace whereas those with a complete sacral fracture and bilateral rami fractures displace at a significantly greater rate (68%).1 The sacrum is the conduit of force transfer from the lower appendicular to the axial skeleton. It transfers the caudally directed forces of the upper body into the lower extremities and delivers rostrally directed forces from the lower body into the spine. In an unstable sacral fracture these forces determine the direction of displacement of the fracture. The caudally directed force of the axial skeleton causes the medial portion of the sacral fracture to displace inferiorly, while the rostrally directed force of the appendicular skeleton causes the lateral portion of the fracture to displace superiorly. The typical motion of the sacrum is described as nutation, or a ‘‘rocking motion.’’ This motion is what directs medial fracture segments to displace anteriorly and lateral segments to displace posteriorly. The rotational component of displacement comes from a fracture or sprain of the anterior ring. In the intact pelvis, the two pubic bones oppose each other and counteract the opposing forces placed on them. With an anterior ring disruption, these forces go unchecked and the affected hemipelvis can rotate around the central axis of the sacrum. In addition numerous force vectors are placed on the fracture segments by muscles that attach to the pelvis. Poor outcomes have been noted with non-operative treatment of sacral fractures.6 In their study of the long-term prognosis of posterior injuries in high-energy pelvic disruption, Dujardin et al. found that sacral fractures had a 75% incidence of unacceptable results in traction and that the primary late complaints were radicular pain rather than pain at the fracture site.7 This failure of non operative management of sacral fractures has led to increased reliance on operative techniques. On the other hand, Siebler et al. reported on nonoperative treatment of Denis zone III sacral fractures and found that it yields consistent healing. However, despite improvement in initial neurologic deficits, residual complaints were common.8 Lindahl and Hirvensalo assert both a correlation between excellent reduction (<5 mm displacement) and outcome as well as the late displacement of minimally displaced sacral fractures and re-displacement of unstable sacral fractures stabilized with iliosacral screws. They recommended stabilization of displaced anterior ring injuries combined with careful attention to the surgical management of unstable sacral fractures.4 Given the generally stable nature of zone I fractures, we hypothesize that a review of cases at our institution will demonstrate zone II fractures to have more displacement at final follow-up than zone I fractures. We also hypothesize that, in nonisolated sacral fractures, patients with anterior stabilization of the pelvic ring with an implant would have less displacement at their fracture sites than patients with either external fixation or no anterior ring stabilization.
Fig. 1. Measurement of vertical displacement on an standard anteroposterior radiograph.
measure vertical displacement and AP displacement of the fracture site. Displacement was independently measured by three observers. The observers measured vertical displacement on each radiograph by constructing a vertical reference line through the lower lumber/ upper sacral area. Horizontal lines 90 degrees perpendicular to this centre line are drawn through bony landmarks.5 The absolute value of the distance between the two horizontal lines was measured as the rostral/caudal (RC) displacement (Fig. 1). AP displacement was measured on the inlet pelvic radiograph (Fig. 2). A reference line was drawn through the pelvis, which evenly bisected the sacrum. Measurement lines were then drawn to the ischial spines and the absolute value of the displacement was recorded again. A difference of at least 5 mm displacement when comparing final follow-up films with initial post-operative films was defined as late displacement. The average change in displacement of the sacral fractures was determined from subtracting post-operative displacement from the final follow-up displacement. Positive values indicate an improvement (or lessening) in displacement and a negative value indicates worsening (or increasing) displacement.
Materials and methods Subjects included patients who were admitted with pelvic fractures to our centre between 2001 and 2007and identified by an ICD-9 and CPT code search in the billing records. A chart review was performed to identify patients older than 18 with associated sacral fractures and concomitant superior and inferior pubic rami fractures. A data-collection tool was created to collect specific patient information from the patients’ electronic medical record and digital imaging databases. In addition to demographic information, radiographic data was also gathered to help classify the fractures. Antero-posterior (AP), inlet, and outlet radiographic views were taken pre-operatively, immediately post-operative and at the final follow-up visit. These radiographs were examined to
Fig. 2. Measurement of posterior displacement on a pelvic inlet radiograph.
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
Patients were excluded if they did not have AP, inlet and outlet radiographs at pre-operative, post-operative, and final follow-up or if their radiographs prevented the evaluators from making the requisite measurements.
Table 2 Average f/u displacement in mm (SD). Overall data
c
Statistical analysis
v
Paired t-tests were used to analyze the overall group data. Kruskal–Wallis tests were used to evaluate the changes between the different methods of anterior stabilization and the different zones of fracture. All analysis was done using SAS v9.1 (SAS Institute Inc., Cary, NC).
Zone 1
Zone 2
P-value
6.60 (5.1) 6.05 (8.9)
6.54 (6.4) 6.92 (5.9)
0.74 0.27
Ex. fix
Ant. fix
No fix
P-value
8.40 (7.1) 6.80 (3.4)
6.45 (4.7) 3.36 (2.9)
4.75 (3.8) 12.5 (19.2)
0.75 0.21
5.45 (5.5) 5.25 (3.7)
6.11 (5.8) 7.82 (8.1)
9.17 (9.0) 7.43 (3.8)
0.69 0.65
Combined c 6.37 (6.0) v 7.50 (6.8)
6.30 (5.1) 4.16 (3.3)
7.40 (7.4) 9.27 (11.2)
0.92 0.21
Zone 1 c
v Zone 2 c
v
Results
201
Over the 6 years period of this study, 70 patients presented, to our facility (a level I trauma centre), with sacral fractures of which 48 patients were found to fit the criteria for this study: older than 18 years with zone I or II sacral fractures and having a full set of pre-operative and post-operative radiographs. The cohort of patients comprised 30 males and 18 females, who were, on average 40 years old. A power calculation determined a sample size of 41 to be necessary for a power of 0.9 at an alpha level of 0.05
average fracture displacements at final follow-up in any of the individual zones or in the entire cohort (Table 1). Statistically significant differences were noted in the average changes in displacement in the patients with zone II fractures in the AP plane (P = 0.03) and in the overall cohort in the AP plane (P = 0.02) (Table 2).
Zones I and II
Discussion
There were 28 patients with zone II fractures and 20 patients with zone I fractures. The zone II fractures showed average fracture displacements of 6.5 mm and 6.9 mm in RC and AP, respectively, at final follow-up. Patients with zone I fractures had average fracture displacements of 6.6 mm and 6.1 mm in the RC and AP directions at final follow-up. There were no significant differences between the zone I and zone II sacral fractures (RC P = 0.74, AP P = 0.24). The average changes in fracture displacement seen in the patients with zone I fracture were 0.6 mm and 1.0 mm in the RC and AP directions. The average changes in fracture displacement seen in the patients with zone II fractures were 1.8 mm and 1.5 mm in the RC and AP directions. There were no significant differences between the average changes in zone I and zone II fractures in any direction (RC P = 0.64 and AP P = 0.68) (Table 1).
This retrospective review demonstrates no statistical difference in the magnitude of fracture displacement at final follow up between zone I and II sacral fractures. Neither was the method used to stabilize the anterior pelvic ring associated with a significant difference in fracture displacement at final follow up. The sample size of 48, although limited, is sufficiently well powered to support the statistical findings. Inappropriately treated or untreated sacral fractures are associated with sequelae such as pain, decreased mobility and neurological compromise.3 We believe our findings to be significant because the results demonstrate that stable early fixation can ensure minimal displacement and thus would reasonably be expected to result in a reduction of the morbidity associated with these fractures; this is in line with the finding that surgery for pelvic ring injuries may improve both early and long outcomes.9,10 The literature focuses on testing different methods of posterior fixation in sacral fractures.11 Many of these studies use anterior fixation to increase the stability of pelvic ring fixation. However, controversy remains as to how much of an increase in stability is gained from anterior fixation. Other reports, such as that by Stocks et al., show that anterior fixation, by either plate and screws or external fixation increases the stiffness of the construct and the load to failure.12 This was confirmed by van Zwienen et al. in their study of 12 cadaveric specimens.13 The difference between these studies was that Stocks et al. focused on posterior fixation with two sacral bars whereas van Zwienen et al. concentrated on one versus two sacroiliac (SI) screws. Sagi et al. also studied the effects of anterior fixation using plate and screws with SI screws for posterior fixation.14 They noted that the addition of anterior fixation with plates and screws significantly decreased linear displacement of the unstable pelvis. They went on to conclude that plate and screw anterior fixation in addition to posterior SI screw fixation (regardless of number or position) restores the normal loading response of the hemi-pelvis. Matta, in his description of the indications for anterior fixation, states that posterior fixation provides a sufficiently stable construct, to prevent the fracture from displacing >1 cm and that anterior fixation carries unnecessary risks.15 Our results were much in the same vein. While the progression or regression of the fracture displacement varied somewhat, we found no statistically significant difference between
Plate and screws/external fixation/no fixation Overall, 16 patients (five with zone I and eleven zone II fractures) had anterior fixation of their pelvic ring with an external fixator, 21 patients (eleven with zone I and ten with zone II fractures) had anterior fixation with plate and screws, and 11 patients (four with zone I and seven zone II) had no anterior fixation. We found no statistically significant differences in the Table 1 Average change in mm (SD). Overall data Zone 1
Zone 2
0.55 (6.3) 1.00 (4.4)
c
v Ex. fix Zone 1 c
v Zone 2 c
v Combined c
v
P-value
1.79 (7.3) 1.53 (7.2) Ant. fix
No fix
0.64 0.68 P-value
0.20 (8.1) 3.20 (3.6)
0.45 (5.8) 0.09 (5.2)
4.25 (5.7) 1.25 (1.9)
0.39 0.37
0.82 (7.0) 4.82 (7.5)
4.50 (8.6) 3.30 (5.6)
0.57 (5.3) 3.29 (5.3)
0.43 0.03
0.50 (7.1) 4.31 (6.4)
1.90 (7.5) 1.62 (5.5)
1.18 (5.7) 2.55 (4.4)
0.96 0.02
202
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
zone I and zone II sacral fractures in either final fracture displacement or average change in fracture displacement from the postoperative period to final follow up, confirming the role of stable early fixation. Our data also showed that the change of fracture displacement from the postoperative period to final follow up was significant (zone II P = 0.03, overall cohort P = 0.02). In both the overall cohort and patients with zone II fractures, anterior stabilization of the pelvic ring with plate and screws yielded positive numbers, indicating a decrease in displacement at the fracture site. Conversely, the patients in these two groups who had either external fixators placed or no fixation of their anterior ring had negative average changes in their fracture displacement, indicating that the fracture displacement increased from the postoperative period to the final follow-up period. Normally described as displacement of >1 cm at final follow up, failure of the mode of stabilization is the focus of much of the published literature on anterior fixation used with posterior stabilization. We found that the mode of anterior fixation played no statistically significant role in how much displacement was noted on the final follow-up radiographs. This would suggest that the instability in zone II fractures may be implicated in the displacement observed over time at the fracture site, while the final displacements in the overall cohorts were not significantly different between the different zones or by the different modes of anterior fixation. A review of the mean values may suggest a trend towards statistical significance when some comparisons are made; examples include zone I ventral displacement of 6.8 mm and 3.36 mm with an external fixator and anterior ring fixation, respectively. This trend was slightly reversed in zone II fractures, where external fixation was associated with follow-up ventral displacement of 5.25 mm, and no fixation was associated with displacement of 7.43 mm. Of note, RC displacement is controlled by the posterior fixation, which mimics the innate mode of stabilization: posterior SI ligaments. All patients in all cohorts averaged <1 cm of displacement at their sacral fracture site at final follow up. The study has limitations, including the heterogeneous population and potential inter-observer variability for measurements. Its strengths lie in the fact that the population, although heterogeneous, was strictly defined. Also, all the data were collected from a single trauma centre, ensuring a degree of uniformity in the protocol for managing these patients’ conditions. These results may have clinical relevance in assessing operative stabilization. If the patient has a small amount of pre-operative displacement, or the displacement can be reduced to nearanatomic size, using external fixation or no fixation are viable options for anterior ring stabilization. However, if the displacement is large or reduction of the fracture will be difficult and result in residual postoperative fracture displacement, anterior fixation with plates and screws should be considered to prevent further displacement or, possibly, to reduce the fracture displacement. A disruption of the anterior ring that was associated with instability in the AP plane was the main consideration prompting anterior fixation. Although there is a theoretic possibility of a fixation involving anterior fixation alone, this method of fixation
was not used in our patient cohort and although not considered a priori, there may be an element of case selection influencing this. If patients have unstable pelvic ring injuries in which anatomic reduction may not be possible, they may nonetheless benefit from anterior fixation with a plate and screws because it may help prevent further AP displacement of the fracture. Conclusions Although optimal treatment for sacral fractures continues to evolve, operative treatment of displaced fractures must be considered; while recent work by Bruce and colleagues elegantly demonstrates the role of non-operative management in some patients,1 the current standard of care would suggest that operative fixation is indicated in unstable fractures. Our cohort comprises patients with zone I and II fractures treated with both conservative and operative approaches. The data presented here demonstrate that operative stabilization ensures that displacement is no worse at follow up than in stable fractures managed more conservatively. Conflict of interest statement The authors have no conflicts of interest to report. References 1. Bruce B, Reilly M, Sims S. OTA highlight paper predicting future displacement of nonoperatively managed lateral compression sacral fractures: can it be done? Journal of Orthopaedic Trauma 2011;25:523–7. 2. Denis F, Davis S, Comfort T. Sacral fractures: an important problem. Retrospective analysis of 236 cases. Clinical Orthopaedics and Related Research 1988;227:67–81. 3. Mehta S, Auerbach JD, Born CT, Chin KR. Sacral fractures. Journal of the American Academy of Orthopaedic Surgeons 2006;14:656–65. 4. Lindahl J, Hirvensalo E. Outcome of operatively treated type-C injuries of the pelvic ring. Acta Orthopaedica 2005;76:667–78. 5. Henderson RC. The long-term results of nonoperatively treated major pelvic disruptions. Journal of Orthopaedic Trauma 1989;3:41–7. 6. Zelle BA, Gruen GS, Hunt T, Speth SR. Sacral fractures with neurological injury: is early decompression beneficial? International Orthopaedics 2004;28:244–51. 7. Dujardin FH, Hossenbaccus M, Duparc F, Biga N, Thomine JM. Long-term functional prognosis of posterior injuries in high-energy pelvic disruption. Journal of Orthopaedic Trauma 1998;12:145–50. 8. Siebler JC, Hasley BP, Mormino MA. Functional outcomes of Denis zone III sacral fractures treated nonoperatively. Journal of Orthopaedic Trauma 2010;24:297– 302. 9. To¨tterman A, Glott T, Madsen JE, Røise O. Unstable sacral fractures: associated injuries and morbidity at 1 year. Spine (Philadelphia PA 1976) 2006;31:E628–35. 10. Routt Jr ML, Simonian PT. Internal fixation of pelvic ring disruptions. Injury 1996;27(Suppl 2):B20–30. 11. Schildhauer TA, Bellabarba C, Nork SE, Barei DP, Routt Jr ML, Chapman JR. Decompression and lumbopelvic fixation for sacral fracture-dislocations with spino-pelvic dissociation. Journal of Orthopaedic Trauma 2006:447–57. 12. Stocks GW, Gabel GT, Noble PC, Hanson GW, Tullos HS. Anterior and posterior internal fixation of vertical shear fractures of the pelvis. Journal of Orthopaedic Research 1991;9:237–45. 13. van Zwienen CM, van den Bosch EW, Snijders CJ, Kleinrensink GJ, van Vugt AB. Biomechanical comparison of sacroiliac screw techniques for unstable pelvic ring fractures. Journal of Orthopaedic Trauma 2004;18:589–95. 14. Sagi HC, Ordway NR, DiPasquale T. Biomechanical analysis of fixation for vertically unstable sacroiliac dislocations with iliosacral screws and symphyseal plating. Journal of Orthopaedic Trauma 2004;18:138–43. 15. Matta JM, Tornetta 3rd P. Internal fixation of unstable pelvic ring injuries. Clinical Orthopaedics and Related Research 1996;329:129–40.
Contents lists available at SciVerse ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
The effect of early operative stabilization on late displacement of zone I and II sacral fractures Osa Emohare a,b,c,*, Nathaniel Slinkard b,c, Paul Lafferty b,c, Sandy Vang b,c, Robert Morgan a,b,c a
Center for Spine and Spinal Cord Injury, Regions Hospital, St Paul, MN, United States Department of Orthopedic Surgery, University of Minnesota, Minneapolis, MN, United States c Department of Orthopedics, Regions Hospital, St Paul, MN, United States b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 4 November 2012
Introduction: This study was designed to evaluate the effect on displacement of early operative stabilization on unstable fractures when compared to stable fractures of the sacrum. Methods: Patient consisted of those sustaining traumatic pelvic fractures that also included sacral fractures of Denis type I and type II classification, who were over 18 at the time of the study. Patients were managed emergently, as judged appropriate at the time and then subsequently divided into two cohorts, comprising those who were either treated operatively or non-operatively. The operative group comprised those treated with either internal fixation or external fixation. Results: Twenty-eight patients had zone II fractures, and 20 had zone I fractures. Zone II fractures showed average displacements of 6.5 mm and 6.9 mm in the rostral–caudal and anteroposterior directions, respectively, at final follow up. Zone I fractures had average displacements of 6.6 mm and 6.1 mm in both directions. There were no significant differences between zone I and II sacral fractures (rostral–caudal P = 0.74, anteroposterior P = 0.24). Average changes in fracture displacement in patients with zone I fractures were 0.6–1.0 mm in both directions. Average changes in zone II fractures were 1.8– 1.5 mm in both directions. There were no significant differences between the average changes in zone I and II fractures in any direction (rostral–caudal P = 0.64, anteroposterior P = 0.68) or in average displacements at final follow up in any of zone or the entire cohort. Statistically significant differences were noted in average changes in displacement in zone II fractures in the anteroposterior plane (P = 0.03) and the overall cohort in the anteroposterior plane (P = 0.02). Conclusion: Operative fixation for unstable sacral fractures ensures displacement at follow up is comparable with stable fractures treated non operatively. ß 2012 Elsevier Ltd. All rights reserved.
Keywords: Fracture fixation Fractures Malunion Orthopaedics Pelvis Sacrum Sacral fractures Denis classification Operative management of sacral fractures Non operative management of sacral fractures
Introduction The optimal approach for the management of sacral fractures has evolved from being predominantly non operative to current practice, where operative intervention is often preferred for displaced fractures.1 Originally reported by Malgaigne in 1876, sacral fractures comprise up to 45% of all pelvic fractures and are commonly caused by high-energy injuries. In a study of 776 pelvic fractures, 236 sacral fractures were found to be present in association with pelvic fractures.2 With an associated pelvic fracture, sacral fractures can often be missed, with up to 50% having been missed historically.3 Although advances in imaging are now improving this rate, significant proportions are still identified late.3
* Corresponding author at: Department of Orthopedics, Regions Hospital, St Paul, MN, United States. Tel.: +1 651 447 1941. E-mail address: [email protected] (O. Emohare). 0020–1383/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.injury.2012.11.004
Three distinct patterns of sacral fractures have been observed and these form the basis for classifying fractures according to the Denis system of classification, as each group has unique characteristics.2 The most commonly observed fracture type is the zone I fracture, which occurs lateral to the sacral foramen; these fractures are more often stable and generally, but not exclusively, treated non operatively. The second most common fracture type is classed as the zone II fracture, which extends through the sacral foramen. Zone II fractures are less stable than zone I fractures and consequently are associated with an increased rate of mal-unions. The third type is the zone III fracture, which is found medial to the sacral foramen and characterized by the presence of concomitant neurological injury in most cases. Studies have shown that patients with fracture displacements >0.5–1 cm have been associated with a higher incidence of non-union, malunion, and long term post operative back pain.4,5 Sacral fractures that initially appear minimally displaced have recently been shown to be at significant risk for late displacement. Bruce et al. found that incomplete lateral compression sacral
200
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
fractures associated with ipsilateral rami fractures can be treated nonoperatively and are unlikely to displace whereas those with a complete sacral fracture and bilateral rami fractures displace at a significantly greater rate (68%).1 The sacrum is the conduit of force transfer from the lower appendicular to the axial skeleton. It transfers the caudally directed forces of the upper body into the lower extremities and delivers rostrally directed forces from the lower body into the spine. In an unstable sacral fracture these forces determine the direction of displacement of the fracture. The caudally directed force of the axial skeleton causes the medial portion of the sacral fracture to displace inferiorly, while the rostrally directed force of the appendicular skeleton causes the lateral portion of the fracture to displace superiorly. The typical motion of the sacrum is described as nutation, or a ‘‘rocking motion.’’ This motion is what directs medial fracture segments to displace anteriorly and lateral segments to displace posteriorly. The rotational component of displacement comes from a fracture or sprain of the anterior ring. In the intact pelvis, the two pubic bones oppose each other and counteract the opposing forces placed on them. With an anterior ring disruption, these forces go unchecked and the affected hemipelvis can rotate around the central axis of the sacrum. In addition numerous force vectors are placed on the fracture segments by muscles that attach to the pelvis. Poor outcomes have been noted with non-operative treatment of sacral fractures.6 In their study of the long-term prognosis of posterior injuries in high-energy pelvic disruption, Dujardin et al. found that sacral fractures had a 75% incidence of unacceptable results in traction and that the primary late complaints were radicular pain rather than pain at the fracture site.7 This failure of non operative management of sacral fractures has led to increased reliance on operative techniques. On the other hand, Siebler et al. reported on nonoperative treatment of Denis zone III sacral fractures and found that it yields consistent healing. However, despite improvement in initial neurologic deficits, residual complaints were common.8 Lindahl and Hirvensalo assert both a correlation between excellent reduction (<5 mm displacement) and outcome as well as the late displacement of minimally displaced sacral fractures and re-displacement of unstable sacral fractures stabilized with iliosacral screws. They recommended stabilization of displaced anterior ring injuries combined with careful attention to the surgical management of unstable sacral fractures.4 Given the generally stable nature of zone I fractures, we hypothesize that a review of cases at our institution will demonstrate zone II fractures to have more displacement at final follow-up than zone I fractures. We also hypothesize that, in nonisolated sacral fractures, patients with anterior stabilization of the pelvic ring with an implant would have less displacement at their fracture sites than patients with either external fixation or no anterior ring stabilization.
Fig. 1. Measurement of vertical displacement on an standard anteroposterior radiograph.
measure vertical displacement and AP displacement of the fracture site. Displacement was independently measured by three observers. The observers measured vertical displacement on each radiograph by constructing a vertical reference line through the lower lumber/ upper sacral area. Horizontal lines 90 degrees perpendicular to this centre line are drawn through bony landmarks.5 The absolute value of the distance between the two horizontal lines was measured as the rostral/caudal (RC) displacement (Fig. 1). AP displacement was measured on the inlet pelvic radiograph (Fig. 2). A reference line was drawn through the pelvis, which evenly bisected the sacrum. Measurement lines were then drawn to the ischial spines and the absolute value of the displacement was recorded again. A difference of at least 5 mm displacement when comparing final follow-up films with initial post-operative films was defined as late displacement. The average change in displacement of the sacral fractures was determined from subtracting post-operative displacement from the final follow-up displacement. Positive values indicate an improvement (or lessening) in displacement and a negative value indicates worsening (or increasing) displacement.
Materials and methods Subjects included patients who were admitted with pelvic fractures to our centre between 2001 and 2007and identified by an ICD-9 and CPT code search in the billing records. A chart review was performed to identify patients older than 18 with associated sacral fractures and concomitant superior and inferior pubic rami fractures. A data-collection tool was created to collect specific patient information from the patients’ electronic medical record and digital imaging databases. In addition to demographic information, radiographic data was also gathered to help classify the fractures. Antero-posterior (AP), inlet, and outlet radiographic views were taken pre-operatively, immediately post-operative and at the final follow-up visit. These radiographs were examined to
Fig. 2. Measurement of posterior displacement on a pelvic inlet radiograph.
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
Patients were excluded if they did not have AP, inlet and outlet radiographs at pre-operative, post-operative, and final follow-up or if their radiographs prevented the evaluators from making the requisite measurements.
Table 2 Average f/u displacement in mm (SD). Overall data
c
Statistical analysis
v
Paired t-tests were used to analyze the overall group data. Kruskal–Wallis tests were used to evaluate the changes between the different methods of anterior stabilization and the different zones of fracture. All analysis was done using SAS v9.1 (SAS Institute Inc., Cary, NC).
Zone 1
Zone 2
P-value
6.60 (5.1) 6.05 (8.9)
6.54 (6.4) 6.92 (5.9)
0.74 0.27
Ex. fix
Ant. fix
No fix
P-value
8.40 (7.1) 6.80 (3.4)
6.45 (4.7) 3.36 (2.9)
4.75 (3.8) 12.5 (19.2)
0.75 0.21
5.45 (5.5) 5.25 (3.7)
6.11 (5.8) 7.82 (8.1)
9.17 (9.0) 7.43 (3.8)
0.69 0.65
Combined c 6.37 (6.0) v 7.50 (6.8)
6.30 (5.1) 4.16 (3.3)
7.40 (7.4) 9.27 (11.2)
0.92 0.21
Zone 1 c
v Zone 2 c
v
Results
201
Over the 6 years period of this study, 70 patients presented, to our facility (a level I trauma centre), with sacral fractures of which 48 patients were found to fit the criteria for this study: older than 18 years with zone I or II sacral fractures and having a full set of pre-operative and post-operative radiographs. The cohort of patients comprised 30 males and 18 females, who were, on average 40 years old. A power calculation determined a sample size of 41 to be necessary for a power of 0.9 at an alpha level of 0.05
average fracture displacements at final follow-up in any of the individual zones or in the entire cohort (Table 1). Statistically significant differences were noted in the average changes in displacement in the patients with zone II fractures in the AP plane (P = 0.03) and in the overall cohort in the AP plane (P = 0.02) (Table 2).
Zones I and II
Discussion
There were 28 patients with zone II fractures and 20 patients with zone I fractures. The zone II fractures showed average fracture displacements of 6.5 mm and 6.9 mm in RC and AP, respectively, at final follow-up. Patients with zone I fractures had average fracture displacements of 6.6 mm and 6.1 mm in the RC and AP directions at final follow-up. There were no significant differences between the zone I and zone II sacral fractures (RC P = 0.74, AP P = 0.24). The average changes in fracture displacement seen in the patients with zone I fracture were 0.6 mm and 1.0 mm in the RC and AP directions. The average changes in fracture displacement seen in the patients with zone II fractures were 1.8 mm and 1.5 mm in the RC and AP directions. There were no significant differences between the average changes in zone I and zone II fractures in any direction (RC P = 0.64 and AP P = 0.68) (Table 1).
This retrospective review demonstrates no statistical difference in the magnitude of fracture displacement at final follow up between zone I and II sacral fractures. Neither was the method used to stabilize the anterior pelvic ring associated with a significant difference in fracture displacement at final follow up. The sample size of 48, although limited, is sufficiently well powered to support the statistical findings. Inappropriately treated or untreated sacral fractures are associated with sequelae such as pain, decreased mobility and neurological compromise.3 We believe our findings to be significant because the results demonstrate that stable early fixation can ensure minimal displacement and thus would reasonably be expected to result in a reduction of the morbidity associated with these fractures; this is in line with the finding that surgery for pelvic ring injuries may improve both early and long outcomes.9,10 The literature focuses on testing different methods of posterior fixation in sacral fractures.11 Many of these studies use anterior fixation to increase the stability of pelvic ring fixation. However, controversy remains as to how much of an increase in stability is gained from anterior fixation. Other reports, such as that by Stocks et al., show that anterior fixation, by either plate and screws or external fixation increases the stiffness of the construct and the load to failure.12 This was confirmed by van Zwienen et al. in their study of 12 cadaveric specimens.13 The difference between these studies was that Stocks et al. focused on posterior fixation with two sacral bars whereas van Zwienen et al. concentrated on one versus two sacroiliac (SI) screws. Sagi et al. also studied the effects of anterior fixation using plate and screws with SI screws for posterior fixation.14 They noted that the addition of anterior fixation with plates and screws significantly decreased linear displacement of the unstable pelvis. They went on to conclude that plate and screw anterior fixation in addition to posterior SI screw fixation (regardless of number or position) restores the normal loading response of the hemi-pelvis. Matta, in his description of the indications for anterior fixation, states that posterior fixation provides a sufficiently stable construct, to prevent the fracture from displacing >1 cm and that anterior fixation carries unnecessary risks.15 Our results were much in the same vein. While the progression or regression of the fracture displacement varied somewhat, we found no statistically significant difference between
Plate and screws/external fixation/no fixation Overall, 16 patients (five with zone I and eleven zone II fractures) had anterior fixation of their pelvic ring with an external fixator, 21 patients (eleven with zone I and ten with zone II fractures) had anterior fixation with plate and screws, and 11 patients (four with zone I and seven zone II) had no anterior fixation. We found no statistically significant differences in the Table 1 Average change in mm (SD). Overall data Zone 1
Zone 2
0.55 (6.3) 1.00 (4.4)
c
v Ex. fix Zone 1 c
v Zone 2 c
v Combined c
v
P-value
1.79 (7.3) 1.53 (7.2) Ant. fix
No fix
0.64 0.68 P-value
0.20 (8.1) 3.20 (3.6)
0.45 (5.8) 0.09 (5.2)
4.25 (5.7) 1.25 (1.9)
0.39 0.37
0.82 (7.0) 4.82 (7.5)
4.50 (8.6) 3.30 (5.6)
0.57 (5.3) 3.29 (5.3)
0.43 0.03
0.50 (7.1) 4.31 (6.4)
1.90 (7.5) 1.62 (5.5)
1.18 (5.7) 2.55 (4.4)
0.96 0.02
202
O. Emohare et al. / Injury, Int. J. Care Injured 44 (2013) 199–202
zone I and zone II sacral fractures in either final fracture displacement or average change in fracture displacement from the postoperative period to final follow up, confirming the role of stable early fixation. Our data also showed that the change of fracture displacement from the postoperative period to final follow up was significant (zone II P = 0.03, overall cohort P = 0.02). In both the overall cohort and patients with zone II fractures, anterior stabilization of the pelvic ring with plate and screws yielded positive numbers, indicating a decrease in displacement at the fracture site. Conversely, the patients in these two groups who had either external fixators placed or no fixation of their anterior ring had negative average changes in their fracture displacement, indicating that the fracture displacement increased from the postoperative period to the final follow-up period. Normally described as displacement of >1 cm at final follow up, failure of the mode of stabilization is the focus of much of the published literature on anterior fixation used with posterior stabilization. We found that the mode of anterior fixation played no statistically significant role in how much displacement was noted on the final follow-up radiographs. This would suggest that the instability in zone II fractures may be implicated in the displacement observed over time at the fracture site, while the final displacements in the overall cohorts were not significantly different between the different zones or by the different modes of anterior fixation. A review of the mean values may suggest a trend towards statistical significance when some comparisons are made; examples include zone I ventral displacement of 6.8 mm and 3.36 mm with an external fixator and anterior ring fixation, respectively. This trend was slightly reversed in zone II fractures, where external fixation was associated with follow-up ventral displacement of 5.25 mm, and no fixation was associated with displacement of 7.43 mm. Of note, RC displacement is controlled by the posterior fixation, which mimics the innate mode of stabilization: posterior SI ligaments. All patients in all cohorts averaged <1 cm of displacement at their sacral fracture site at final follow up. The study has limitations, including the heterogeneous population and potential inter-observer variability for measurements. Its strengths lie in the fact that the population, although heterogeneous, was strictly defined. Also, all the data were collected from a single trauma centre, ensuring a degree of uniformity in the protocol for managing these patients’ conditions. These results may have clinical relevance in assessing operative stabilization. If the patient has a small amount of pre-operative displacement, or the displacement can be reduced to nearanatomic size, using external fixation or no fixation are viable options for anterior ring stabilization. However, if the displacement is large or reduction of the fracture will be difficult and result in residual postoperative fracture displacement, anterior fixation with plates and screws should be considered to prevent further displacement or, possibly, to reduce the fracture displacement. A disruption of the anterior ring that was associated with instability in the AP plane was the main consideration prompting anterior fixation. Although there is a theoretic possibility of a fixation involving anterior fixation alone, this method of fixation
was not used in our patient cohort and although not considered a priori, there may be an element of case selection influencing this. If patients have unstable pelvic ring injuries in which anatomic reduction may not be possible, they may nonetheless benefit from anterior fixation with a plate and screws because it may help prevent further AP displacement of the fracture. Conclusions Although optimal treatment for sacral fractures continues to evolve, operative treatment of displaced fractures must be considered; while recent work by Bruce and colleagues elegantly demonstrates the role of non-operative management in some patients,1 the current standard of care would suggest that operative fixation is indicated in unstable fractures. Our cohort comprises patients with zone I and II fractures treated with both conservative and operative approaches. The data presented here demonstrate that operative stabilization ensures that displacement is no worse at follow up than in stable fractures managed more conservatively. Conflict of interest statement The authors have no conflicts of interest to report. References 1. Bruce B, Reilly M, Sims S. OTA highlight paper predicting future displacement of nonoperatively managed lateral compression sacral fractures: can it be done? Journal of Orthopaedic Trauma 2011;25:523–7. 2. Denis F, Davis S, Comfort T. Sacral fractures: an important problem. Retrospective analysis of 236 cases. Clinical Orthopaedics and Related Research 1988;227:67–81. 3. Mehta S, Auerbach JD, Born CT, Chin KR. Sacral fractures. Journal of the American Academy of Orthopaedic Surgeons 2006;14:656–65. 4. Lindahl J, Hirvensalo E. Outcome of operatively treated type-C injuries of the pelvic ring. Acta Orthopaedica 2005;76:667–78. 5. Henderson RC. The long-term results of nonoperatively treated major pelvic disruptions. Journal of Orthopaedic Trauma 1989;3:41–7. 6. Zelle BA, Gruen GS, Hunt T, Speth SR. Sacral fractures with neurological injury: is early decompression beneficial? International Orthopaedics 2004;28:244–51. 7. Dujardin FH, Hossenbaccus M, Duparc F, Biga N, Thomine JM. Long-term functional prognosis of posterior injuries in high-energy pelvic disruption. Journal of Orthopaedic Trauma 1998;12:145–50. 8. Siebler JC, Hasley BP, Mormino MA. Functional outcomes of Denis zone III sacral fractures treated nonoperatively. Journal of Orthopaedic Trauma 2010;24:297– 302. 9. To¨tterman A, Glott T, Madsen JE, Røise O. Unstable sacral fractures: associated injuries and morbidity at 1 year. Spine (Philadelphia PA 1976) 2006;31:E628–35. 10. Routt Jr ML, Simonian PT. Internal fixation of pelvic ring disruptions. Injury 1996;27(Suppl 2):B20–30. 11. Schildhauer TA, Bellabarba C, Nork SE, Barei DP, Routt Jr ML, Chapman JR. Decompression and lumbopelvic fixation for sacral fracture-dislocations with spino-pelvic dissociation. Journal of Orthopaedic Trauma 2006:447–57. 12. Stocks GW, Gabel GT, Noble PC, Hanson GW, Tullos HS. Anterior and posterior internal fixation of vertical shear fractures of the pelvis. Journal of Orthopaedic Research 1991;9:237–45. 13. van Zwienen CM, van den Bosch EW, Snijders CJ, Kleinrensink GJ, van Vugt AB. Biomechanical comparison of sacroiliac screw techniques for unstable pelvic ring fractures. Journal of Orthopaedic Trauma 2004;18:589–95. 14. Sagi HC, Ordway NR, DiPasquale T. Biomechanical analysis of fixation for vertically unstable sacroiliac dislocations with iliosacral screws and symphyseal plating. Journal of Orthopaedic Trauma 2004;18:138–43. 15. Matta JM, Tornetta 3rd P. Internal fixation of unstable pelvic ring injuries. Clinical Orthopaedics and Related Research 1996;329:129–40.