ORIGINAL ARTICLE

Functional Outcomes of Denis Zone III Sacral Fractures Treated Nonoperatively

Justin C. Siebler, MD, Brian P. Hasley, MD, and Matthew A. Mormino, MD

Objectives: The purpose of this study is to report the outcomes of

nonoperative treatment in patients with Denis Zone III sacral fractures at a minimum of 2 years follow up.

INTRODUCTION
High-energy pelvis fractures are often the result of motor vehicle collisions or falls from a height and frequently include sacral fractures. Sacral fractures that cross the midline result in spinopelvic dissociation and are the least common and the least studied of sacral fractures. These are associated with higher rates of neurologic injury, including sensory and motor decits in the lower extremities, saddle anesthesia, and bowel, bladder, and sexual dysfunction.1 The most often cited classications of sacral fractures are the Denis1 (Fig. 1), Roy-Camille2 (with Strange-Vognsen modication)3 (Fig. 1), and an alphabet description4 (Fig. 2). Unfortunately, displacement and comminution can be quite variable and this has led to the lack of consistent reporting of fracture morphology and classication. The studies of nonoperative treatment of Denis Zone III sacral fractures are limited to small case reports or subgroups of larger series with limited clinical outcome measures and without validated functional outcomes.13,59 Historically, nonoperative treatment has been the standard care for these fractures unless accompanied by other pelvic ring instability. Recent literature has been published proposing surgical treatment of these fractures.1016 However, there are no clear published guidelines as to the surgical indications for this rare fracture pattern. The purpose of this study is to present our cohort of patients with a minimum of 2 years follow up of nonoperatively treated Zone III sacral fractures and their functional outcomes.

Design: Retrospective review of prospectively collected data of

a consecutive series of patients.

Setting: Level I trauma center. Patients: A consecutive series of 15 patients (1547 years old) with Denis Zone III sacral fractures treated nonoperatively from 1997 to 2002 was studied. Eleven patients were available for follow-up questionnaires; nine participated in a physical examination. Time to nal follow up averaged 43 months (range, 2567 months). Intervention: Demographic data; mechanism of injury; injuryspecic assessment of bowel, bladder, and sexual function; physical examination; and fracture pattern were collected from a prospectively collected database. Main Outcome Measurements: At a minimum of 2-year follow up, evaluation of SF-36 scores, Roland Morris back pain questionnaire, and Gibbons classication was conducted. Results: All fractures healed. Six patients had a postinjury increase in kyphosis (range, 117) without a correlation to nal outcomes. Mean SF-36 scores were all uniformly lower than the normalized general population and were biased by frequent associated injuries. Final Roland-Morris scores averaged 3.3 6 3.3. Gibbons classication scores initially averaged 2 6 1.2 and decreased to 1.5 6 0.8, each within their standard deviations. Eight had residual bowel, bladder, and/or sexual dysfunction. Conclusions: Nonoperative treatment of Denis Zone III sacral fractures yields consistent healing. Despite improvement in initial neurologic decits, residual complaints were common.
Key Words: sacral fractures, nonoperative treatment (J Orthop Trauma 2010;24:297302)

PATIENTS AND METHODS

Fifteen patients with Denis Zone III sacral fractures treated nonoperatively over a 5-year period (19972002) were culled from a prospectively gathered orthopaedic trauma database stored in an Excel (Microsoft Inc., Redmond, WA) spreadsheet and their data retrospectively reviewed. One died 2 months from the date of injury. Three patients could either not be contacted or refused to participate in follow-up studies. Eleven patients agreed to participate. These patients had no rotational or vertical displacement of their hemipelvis at initial presentation and throughout treatment. The amount of displacement or perceived stability of the Zone III sacral portion of the fracture did not affect inclusion; all were included. No reduction maneuvers were attempted. Three additional patients with Zone III fractures were treated operatively during this time as a result of a vertically and/or
www.jorthotrauma.com |

Accepted for publication November 25, 2009. From the Department of Orthopaedic Surgery and Rehabilitation, University of Nebraska Medical Center, Omaha, NE. None of the authors have any nancial disclosures in regards to this manuscript. There are no devices included in this manuscript that are subject to US Food and Drug Administration approval. Reprints: Justin C. Siebler, MD, Department of Orthopaedic Surgery and Rehabilitation, University of Nebraska Medical Center, 981080 Nebraska Medical Center, Omaha, NE 68198-1080 (e-mail: jsiebler@unmc.edu). Copyright 2010 by Lippincott Williams & Wilkins

J Orthop Trauma  Volume 24, Number 5, May 2010

297

Siebler et al

J Orthop Trauma  Volume 24, Number 5, May 2010

FIGURE 1. (A) Sacral fracture classication as described by Denis.1 Zone 1 is lateral to the neuroforamina, Zone 2 involves the neuroforamina but not the spinal canal, and Zone 3 extends into the spinal canal. (B) Denis Zone III sacral fracture classication as described by Roy-Camille2 and modied by Strange-Vognsen modication.3 Type 1 are anteriorly exed with no translation. Type 2 are anteriorly exed with posterior translation of cephalad fragment. Type 3 have complete translation of fragments. Type 4 has total comminution without displacement of cephalad fragment. (Reprinted with permission from J Am Acad Orthop Surg. 2006;14:656665. 2006 AAOS.)

rotationally unstable pelvic ring injury and were not included in the study. Initial data gathered included demographics, mechanism of injury, fracture pattern/classication, concomitant injuries, neurologic and/or motor decits, and bowel and bladder function. The 11 patients included seven males and four females. Their mean age at time of injury was 28 years (range, 1547 years). Mechanism of injury included eight motor vehicle crashes (four patients were ejected from the vehicle) and three

falls from a height. Concomitant injuries were frequent. Eight patients had thoracic and lumbar spine fractures. Although transverse and spinous process fractures were the most common, two had compression fractures and one had a burst fracture. All spine injuries were stable and treated nonoperatively. The initial radiographic evaluation included anteroposterior, inlet and outlet pelvis views, and lateral sacral views. Axial, coronal, and sagittal computed tomography scans were evaluated (GE HiSpeed Advantage, Waukesha, WI, 1997, 35-mm slice thickness). Fracture patterns were classied as Denis Zone III fractures and further classied using the RoyCamille classication system and by a description of fracture morphology. There were six Roy-Camille Type 2 fracture patterns (Fig. 3) and ve Roy-Camille Type 1 fracture patterns. The radiographic views obtained at follow up were the same as those obtained initially. Initial and nal angulation was measured from lateral sacral radiographs by the method described by Schildhauer et al.10 The kyphotic angle was measured by an angle formed from lines drawn along the posterior cortex of the sacral body above and below the transverse fracture line. Seven of the 11 had initial kyphotic deformity that averaged 57 (range, 4386). The three authors reviewed all radiographs blindly and independently and then a consensus was obtained if differences existed. Mobilization of these patients varied according to fracture pattern, initial angular displacement, associated injuries, and rate of healing. In general, patients with transverse or lambda-type fractures were initially treated with toe-touch weightbearing for 6 to 12 weeks followed by gradual progression of weightbearing. Patients with U- or H-type fractures were initially treated with bed to chair transfers for 6 to 9 weeks followed by progressive weightbearing. Patients were evaluated at 2, 6, 12, 24, and 52 weeks and nal follow up to monitor any changes from their initial examination and assess fracture healing. Patients were asked to return for nal evaluation after a minimum of 2 years from the time of injury. Mean follow up from time of injury was 43 6 14 months (range, 2567 months). Two of the 11 patients refused to have a follow-up physical examination but completed the functional outcome questionnaires. Three English or Spanish questionnaires were administered: the Standardized Short Form-3617,18 with general population-normalized scores (higher scores are better), the Roland-Morris 24-question back pain survey,19,20 which is a self-rated disability form to measure disability resulting from back pain (lower scores are better), and a form with specic questions related to bowel, bladder, and sexual function. Records were reviewed to determine the initial and nal Gibbons classication scores. This assigns a score of 1 to 4 for neurologic decit (Table 1).6 Gibbons classication initially averaged 2 6 1.2. The senior author performed all the initial and follow-up physical examinations. Demographic, examination, and radiographic ndings are presented in Table 2.

RESULTS
All fractures healed; there were no nonunions. Six patients had an increase in kyphosis from their injury radiographs to their nal follow-up radiographs with a range
q 2010 Lippincott Williams & Wilkins

298

| www.jorthotrauma.com

J Orthop Trauma  Volume 24, Number 5, May 2010

Functional Outcomes of Denis Zone III Sacral Fractures

FIGURE 2. Complex Denis Zone III sacral fractures can be classied descriptively by the letter of the alphabet they most closely resemble. From left-to-right and top-to-bottom H, U, l, and T. (Reprinted with permission from J Am Acad Orthop Surg. 2006;14:656665. 2006 AAOS.)

of increase of 1 to 17. There was no correlation of nal degree of kyphosis to any of the outcome measures. Two patients (No. 10 and 11 in Table 2) had a worsening of their symptoms and a concomitant increase in their Gibbons classication, increasing the mean to 2.4 6 1.3. Patient 11 developed increasing lower extremity paresthesias 4 weeks after injury and underwent a sacral decompression with resolution of his paresthesias. Despite this, he has persistent 4/5 strength of foot eversion bilaterally and decreased light

touch sensation in the L5 distribution of his right foot. Patient 10 had transient loss of bowel and bladder function after being noncompliant with his weightbearing restrictions. His neurologic status improved within 24 hours of resuming bedto-chair transfers and had full bowel and bladder control at nal evaluation. At nal follow up, the mean Gibbons score improved to 1.5 6 0.8. All three patients (No. 1, 2, and 10 in Table 2) with complete bowel and bladder loss (Gibbons 4) had restoration of voluntary sphincter control at follow up.

FIGURE 3. Anteroposterior pelvis (A) and lateral sacral (B) radiographs demonstrating sacral fracture with intact pelvic ring. Sagittal computed tomographic image (C) of RoyCamille Type 1 sacral fracture. (Patient 1 in Table 1.) q 2010 Lippincott Williams & Wilkins www.jorthotrauma.com |

299

Siebler et al

J Orthop Trauma  Volume 24, Number 5, May 2010

TABLE 1. Gibbons Classification of Neurologic Injury5

Type 1 2 3 4 Neurologic Decit None Paresthesias only Lower extremity motor decit Bowel/bladder dysfunction

Five patients reported urinary frequency as the most common bladder symptom (Table 3). Six reported constipation as a new problem after injury as the most common bowel complaint. Four of these admitted to having uncontrolled bowel movements at times. Three of these with constipation did not have any initial impairment in their bowel function at the time of injury. Four had problems with sexual function after their injury. Seven of 11 have some degree of subjective hypoesthesia or paresthesia in their lower extremities. Averaged normalized SF-36 scores for all 11 responders in the eight health concepts were lower than the general population but within one standard deviation (Fig 4). The scores for the general population have been normalized to a mean of 50 and a standard deviation of 10. The lowest average score of the eight health concepts was physical function at 41 and the highest was body pain at 50. The average physical component summary score was 45 (range, 2859). The mental component summary score average was 46 (range, 2662). Some individual scores fell more than one standard deviation below the mean. Lower back/sacrum pain was reported by almost all patients in varying degrees of intensity and periodicity. The Rolland-Morris back pain questionnaire score mean was 3.3 6 3.3 (range, 09 out of 24). The most frequent complaint was the need to change position frequently to improve comfort (six of 11). Physical examination ndings at the nal follow-up examination demonstrated a decrease in light touch sensation

in various distributions in ve of nine patients examined. Two patients had decreased motor strength. One (No. 11, Table 2) had 4/5 eversion of his bilateral ankles. Patient 1 had 4/5 unilateral great toe dorsiexion strength. Eight of nine patients examined had decreased or absent Achilles reex. Two complications occurred. Patient 11 (Table 2) developed progressive lower extremity paresthesias 4 weeks after injury for which he underwent sacral decompression. A 53-year-old woman was ventilator-dependent for 4 weeks from the time of admission and her injuries included a H-type sacral fracture, L1 and L3 burst fractures (treated nonoperatively), bilateral pneumothoraces, and a liver contusion. After she regained consciousness, she was treated with bed-to-chair mobilization. She died approximately 2 months from the date of injury of a pulmonary embolism.

DISCUSSION
In the current study, we present the 2-year outcomes of nonoperatively treated Denis Zone III sacral fractures. Although the majority of our patients had residual symptoms, their SF-36 scores are lower but within one standard deviation of the reported normal population. All fractures healed; some demonstrated an increase in kyphosis from initial injury to nal healed position. This increase was not correlated with abnormal physical examination ndings or difference in responses to the questionnaires administered. In only one patient (No. 11, Table 2) was a sacral decompression performed for worsening neurologic symptoms. Despite this, neurologic symptoms improved for all other patients as demonstrated by their Gibbons scores. However, most of the patients still had residual neurologic symptoms and pain, the most common being sacral and buttock pain and paresthesias or hypoesthesia in the back, buttock, or lower extremities. Constipation was the most common bowel complaint, whereas urinary frequency was the most common urinary tract complaint. The bowel, bladder, and

TABLE 2. Study Population Demographic Data

Patient Age at No. Sex DOI 1 2 3 4 5 6 7 8 9 10 11 M F M F F M F M M M M 45 37 21 15 21 33 17 22 27 47 19 Fracture Type Initial Final Kyphotic Kyphotic Gibbons Gibbons Angle Angle R-C Type Initial Final 43 48 0 0 0 86 0 73 47 60 43 46 58 0 0 0 84 0 88 48 77 44 2 2 1 1 1 2 1 1 2 2 2 4 4 1 1 3 1 1 2 2 2 then 4 1 then 3 3 2 1 1 2 1 1 1 1 1 3 Roland Follow-Up Morris in Months 9 0 2 0 8 7 1 2 4 1 2 56 67 38 51 63 44 39 28 31 36 25

MOI

Treatment in Weeks BC 3 8 BC 3 6, then transfers only 3 3 BC 3 6, then transfers only 3 3, with TLSO TTWB 3 12 TTWB 3 12 WBAT, Jewitt brace 3 12 TTWB 3 6 BC BC BC BC 3 3 3 3 8 8 8 9

Ejected MVC U-Type MVC U-Type Ejected MVC l-Type l-Type l-Type Transverse Transverse, Left alar Ejected MVC H-Type Fall U-Type Fall U-Type Ejected MVC U-Type MVC MVC Fall MVC

DOI, date of injury; MOI, mechanism of injury; R-C, Roy-Camille; M, male; F, female; MVC, motor vehicle crash; BC, bed-to-chair transfers; TTWB, toe-touch weightbearing.

300

| www.jorthotrauma.com

q 2010 Lippincott Williams & Wilkins

J Orthop Trauma  Volume 24, Number 5, May 2010

Functional Outcomes of Denis Zone III Sacral Fractures

TABLE 3. Number of Patients with Findings at Final Follow Up (totals may not add Up, because patients may have more than one finding)
Finding Urinary symptoms Urinary frequency Urinary urgency Nocturia Difculty initiating urination More than one of above Bowel symptoms Constipation Loss of control More than one of the above Sexual symptoms Dysparenuria Difculty with orgasm Erectile dysfunction More than one of above Pain (back, buttock, lower extremities) Sacral prominence Hypo-/paresthesias (back, buttock, lower extremities) Subjective weakness (back, buttock, lower extremities) Physical examination Decreased lower extremity sensation Decreased deep tendon reexes Number of Patients 6 total 5 1 2 1 2 6 total 6 4 2 4 total 1 2 2 3 7 3 7 4 (55%) (45%) (9%) (18%) (9%) (18%) (55%) (55%) (36%) (18%) (36%) (9%) (18%) (18%) (27%) (64%) (27%) (64%) (36%)

5 (56%) 8 (89%)

sexual dysfunction in our study data is consistent with the pelvic fracture data published by Copeland et al,21 albeit their study included a mixed cohort with various types of pelvis fractures not directly comparable to ours. Two of the initial studies to evaluate internal xation of Denis Zone III sacral fractures, with and without decompression, noted 22% (two of nine)11 and 50% (two of four)12 of their patients with initial neurologic decits had residual neurologic decits postoperatively. In 2007, Totterman et al13 reported on the functional outcome of 31 patients treated surgically as a result of displaced sacral fractures. Seven of 31

(22.5%) had Denis Zone III fractures. There was no single method of xation or decompression, because each case was individualized. They reported all components of their SF-36 scores were signicantly lower for their Norwegian population at an average follow up of 15 months. Bowel, bladder, and sexual impairment were common, ranging from 35% to 48% of patients. These three studies involving operative xation describe similar patients with persistent neurologic decits and decreased SF-36 scores at nal evaluation compared with our nonoperatively treated patients. Schildhauer et al10 described their ndings of lumbopelvic xation for sacral fracturedislocations; all patients had Denis Zone III sacral fractures (Roy-Camille 24) and cauda equina (Gibbons 4) at presentation. All were treated with open reduction, sacral decompression, and lumbopelvic xation. They noted a high rate of sacral nerve root transection (11 of 19) at the time of surgery and correlated this with worse outcomes. However, four of these 11 had complete recovery of bowel and bladder function. Kyphosis improved from 43 to 21 on average. Fifteen of 18 (83%) patients had improvement in bowel and bladder function at a minimum of 12 months follow up. Average Gibbons scores improved from 4 to 2.8 at 31-month average follow up. They further discussed their complications associated with this group14 and demonstrated six of 19 (31%) had broken rod(s) and three of 19 (16%) had an infection. A total of eight of 19 (42%) had an unplanned second operation resulting from infection, wound healing disruption, or iliac screw prominence. Our results showed overall neurologic improvement in 10 of our 11 patients. In comparison to Schildhauer et als study, all three of our patients with similar initial loss of bowel and bladder function had complete recovery. When comparing our Gibbons 3 and 4 patients (ve) with their similar patients, our patients showed an improvement of 1.4 Gibbons scores, which is comparable to their 1.2 score improvement. Certainly only limited comparisons can be made between the patient groups because they are not totally equal and are evaluated by separate observers. However, we chose to use similar outcomes measures and classications as these researchers in an attempt to overcome these shortcomings.

FIGURE 4. Normalized SF-36 scores of study population versus general population all study data points fall within 1 SD of normalized data. q 2010 Lippincott Williams & Wilkins www.jorthotrauma.com |

301

Siebler et al

J Orthop Trauma  Volume 24, Number 5, May 2010

2. Roy-Camille R, Saillant G, Gagna G, et al. Transverse fracture of the upper sacrum. Suicidal jumpers fracture. Spine. 1985;10:838845. 3. Strange-Vognsen HH, Lebech A. An unusual type of fracture in the upper sacrum. J Orthop Trauma. 1991;5:200203. 4. Vaccaro AR, Kim DH, Brodke DS, et al. Diagnosis and management of sacral spine fractures. J Bone Joint Surg Am. 2004;86:166175. 5. Phelan ST, Jones DA, Bishay M. Conservative management of transverse fractures of the sacrum with neurological features. A report of four cases. J Bone Joint Surg Br. 1991;73:969971. 6. Gibbons K, Soloniuk D, Razack N. Neurological injury and patterns of sacral fractures. J Neurosurg. 1990;72:889893. 7. Fountain SS, Hamilton RD, Jameson RM. Transverse fractures of the sacrum. A report of six cases. J Bone Joint Surg Am. 1977;59:486489. 8. Bucknill TM, Blackburne JS. Fracturedislocations of the sacrum. Report of three cases. J Bone Joint Surg Br. 1976;58:467470. 9. Ebraheim NA, Lu J, Biyani A, et al. Zone III fractures of the sacrum. A case report. Spine. 1996;21:23902396. 10. Schildhauer TA, Bellabarba C, Nork S, et al. Decompression and lumbopelvic xation for sacral fracturedislocations with spino-pelvic dissociation. J Orthop Trauma. 2006;20:447457. 11. Nork SE, Jones CB, Harding SP, et al. Percutaneous stabilization of U-shaped sacral fractures using iliosacral screws: technique and early results. J Orthop Trauma. 2001;15:238246. 12. Hunt N, Jennings A, Smith M. Current management of U-shaped sacral fractures or spino-pelvic dissociation. Injury. 2002;33:123126. 13. Totterman A, Glott T, Lundgaard Sberg H. et al. Pelvic trauma with displaced sacral fractures functional outcome at one year. Spine. 2007;32: 14371443. 14. Bellabarba C, Schildhauer T, Vaccaro A, et al. Complications associated with surgical stabilization of high-grade sacral fracture dislocations with spino-pelvic instability. Spine. 2006;31;S80S88. 15. Robles LA. Transverse sacral fractures: a review. Spine J. November 5, 2007. [Epub ahead of print]. 16. Pohlemann T, Culemann, U. Summary of controversial debates during the 5th Homburg Pelvic Course 1315 September, 2006. Injury. 2007;38: 424430. 17. Ware J, Kosinski M, Keller S. SF-36 Physical and Mental Health Summary Scales: A Users Manual. Boston, MA: The Heath Institute; 1994. 18. Ware J, Kosinski M. SF-36 Physical and Mental Health Summary Scales: A Manual for Users of Version 1, 2nd ed. Lincoln, RI: QualityMetric Incorporated; 2001. 19. Roland M, Morris R. A study of the natural history of back pain. Part I: development of a reliable and sensitive measure of disability in low-back pain. Spine. 1983;8:141144. 20. Roland M, Fairbank J. The Roland-Morris Disability Questionnaire and the Oswestry Disability Questionnaire. Spine. 2000;25:31153124. 21. Copeland CE, Bosse M, McCarthy M, et al. Effect of trauma and pelvic fracture on female genitourinary, sexual, and reproductive function. J Orthop Trauma. 1997;11:7381. 22. Mehta S, Auerbach JD, Born CT, et al. Sacral fractures. J Am Acad Orthop Surg. 2006;14:656665. 23. Schmidek HH, Smith DA, Kristiansen TK. Sacral fractures. Neurosurgery. 1984;15:735746. 24. Kim MY, Reidy DP, Nolan PC, et al. Transverse sacral fractures: case series and literature review. Can J Surg. 2001;44:359363.

No previous studies or reviews4,15,2124 have demonstrated a difference in outcomes of patients treated operatively versus nonoperatively. The literature would suggest that neurologic improvement could be expected in nearly 80% of patients regardless of operative or nonoperative management.5,15,2224 In a recent review of the literature, Robles15 evaluated 29 articles with 90 patients experiencing transverse sacral fractures treated with both operative (77%) and nonoperative care (23%). Operative care included decompression alone, stabilization alone, or both. He pointed out the difculty in comparing dissimilar patient groups with varied outcome measures and was unable to demonstrate denitive treatment guidelines from his review. Robles and those at the fth Homburg Pelvic Course16 state their preference for surgical management while acknowledging the neurologic outcome is probably more dependent on the injury itself and recovery is frequently seen in nonoperative treatment. The limitations of our study include its retrospective review, no blinding, and no internal control group. These are compounded by the loss of four of 15 patients to follow up. Also, the heterogeneity of these fracture types leads to difculty comparing the patients in our cohort with each other and those in other cohorts. The patient numbers are low; however, we know of no larger study with functional outcome data on nonoperative treatment of this type of injury. Although the purpose of our study was not to advocate treatment, nor was it to identify indications for surgical treatment, it is noteworthy that four of ve U-shaped fracture patterns were Gibbons 3 or 4 and had some of the highest initial kyphotic angulation. It may be this fracture pattern that would most benet from decompression, reduction, and stabilization. Our outcomes data provide a comparison group for similarly classied fractures which surgical management of these fractures can be compared. Clearly, a multicenter study comparing operative and nonoperative treatment of these rare fractures would be ideal to comment on the potential superiority of either method. ACKNOWLEDGMENT We thank Harlan Sayles for his assistance with the statistical analysis in preparing this manuscript. REFERENCES
1. Denis F, Davis S, Comfort T. Sacral fractures: an important problem. Retrospective analysis of 236 cases. Clin Orthop Relat Res. 1988;227: 6781.

302

| www.jorthotrauma.com

q 2010 Lippincott Williams & Wilkins