➢ Spinopelvic fixation is often required to stabilize vertically unstable and complex bilateral sacral fractures.
➢ Biomechanically, triangular osteosynthesis, a combination of spinopelvic fixation along with sacroiliac or transsacral screw fixation, provides the greatest stability in vertically unstable sacral fractures.
➢ Wound-healing and infection rates have been reported as high as 26%, and iatrogenic nerve injuries occur in up to 13% of patients with complex sacral fractures.
Vertically unstable and complex bilateral sacral fractures are a result of high-energy mechanisms that are often associated with other injuries1-19. Vertically unstable sacral fractures result in the separation of the hemipelvis from the lumbar spine, whereas complex bilateral fractures result in a complete dissociation of the appendicular and axial skeleton. Recently, the surgical treatment algorithm for these fractures has changed, with an increasing number of surgeons advocating for the use of spinopelvic fixation or triangular osteosynthesis, a combination of spinopelvic fixation along with sacroiliac or transsacral screw fixation, to increase the stability of the construct and to allow for early weight-bearing with decreased concern for secondary displacement1-19.
Osseous and Ligamentous Anatomy
The sacrum is a rhomboid-shaped bone consisting of five fused segments that is the lowest functional section of the spinal column and acts as the keystone of the pelvic ring20. When force enters the sacrum, it propagates laterally across the sacroiliac joints into the pelvis. Because of the shape of the sacrum and the orientation of the sacroiliac joints, axial forces stabilize the sacroiliac articulation in the pelvic outlet plane; however, in the pelvic inlet plane, the osseous configuration lacks inherent stability. The interosseous ligaments and posterior sacroiliac ligaments provide the majority of the intrinsic stability. Comparatively, the sacrospinous and sacrotuberous ligaments provide the extrinsic stability21.
The osseous architecture of the sacrum can be quite variable, with upper sacral morphologic differences occurring in twenty-eight (35%) of eighty patients22. The identification of sacral dysmorphism is critical in patients with sacral fractures, as misunderstanding of differing sacral abnormalities can lead to malpositioning of iliosacral screws22-26.
Despite the upper sacral variability, the sacrum has the highest and most consistent density of cancellous bone, making the upper sacral body the preferred location for instrumentation27,28. As patients age, bone resorption occurs, most notably in the sacral ala27. Resorption begins underneath the S1 facet and can lead to a large void of trabecular bone encompassing nearly the entire sacral ala in patients with osteoporosis27.
The spinal canal in the sacrum is formed by the posterior aspect of the sacral body anteriorly, the sacral pedicles laterally, and the laminae posteriorly. Moving caudally, the canal decreases in size, changing from a triangular shape at S1 to a narrow and flat canal in the caudal sacrum20. Although the spinal cord typically terminates around L2, the filum terminale continues through the sacrum and anchors the spinal cord to the coccyx. The thecal sac and nerve roots accompany the filum terminale in the upper sacrum; however, the thecal sac terminates near S220.
Four paired ventral and dorsal neuroforamina are located at the confluence of the sacral body and ala. Ventrally, the L4, L5, and sacral nerve roots form the sacral plexus. The L4 nerve travels medial to the psoas along the middle of the sacral ala, and just cephalad to the first sacral foramen and within 1 cm of the sacroiliac joint, it anastomoses with the L5 nerve root. The L5 nerve root travels beneath the inferior edge of the sacrolumbar ligament and passes over the anterosuperior aspect of the sacral ala29,30. Because of the intimate relationship with the sacrum, the L4 and L5 nerve roots are at a high risk of injury29,30.
Additionally, the sympathetic chain drapes the ventral sacrum medial to the neuroforamina and connects with the ventral nerve roots. Because of this connection, the sacral plexus is responsible for bowel, bladder, and sexual function20,30. Fortunately, bilateral innervation of the perineal structures often allows for continued bowel and bladder function with a complete unilateral disruption31.
Classification of Sacral Fractures
Multiple different sacral fracture classifications have been proposed32-35; however, the classification system of Denis et al.32 is the basis for most classifications. Denis et al. proposed classifying sacral fractures based on the fracture’s location in relationship to the neural foramen (Fig. 1). Fractures lateral to the neural foramina are Zone I, fractures through the foramen are Zone II, and fractures medial to the foramen are Zone III32. Aside from the simplicity, the predominant advantage of this classification is its ability to predict neurologic deficits. In 236 patients, Zone-I fractures were the most common fracture reported (118 patients [50%]), and they had the best neurologic outcome, with a neurologic deficit in only 6% (seven of 118 patients)32. Zone-II fractures were the second most common fracture (eighty-one [34%] of 236 patients) and resulted in a neurologic deficit in 28% (twenty-three of eighty-one patients)32. Zone-III fractures only occurred in 16% (thirty-seven of 236 patients), but, of these thirty-seven patients, 57% (twenty-one patients) had a neurologic injury and 43% (sixteen patients) had bowel and bladder dysfunction32.
Expanding on the classification system of Denis et al., Isler further classified Zone-II sacral fractures on the basis of their relationship to the L5-S1 facet joint (Fig. 2)35. Type-A fractures exit lateral to the facet and therefore do not affect spinopelvic stability, whereas Type-B fractures exit through the L5-S1 facet, and Type-C fractures exit medial to the L5-S1 facet. Type-B and Type-C fractures are vertically unstable fractures that can lead to spinopelvic instability35.
In another enhancement of the classification of Denis et al., Roy-Camille et al. proposed a subclassification for bilateral Zone-III sacral fractures with a transverse component leading to a lumbopelvic dissociation (U-type sacral fractures) (Fig. 3)33. Type-1 fractures have kyphotic angulation across the transverse fracture line without translation. Type-2 fractures have kyphosis with <100% translation, and type-3 fractures have kyphosis with >100% translation33. Although not in the original classification, type 4, a fracture with minimal displacement but substantial comminution of the cephalad segment, was added by Strange-Vognsen and Lebech34. Lastly, fractures that involve both sides of the sacrum can be described on the basis of which letter of the alphabet the fracture most closely resembles (i.e., U type, H type, T type, or lambda type)36 (Fig. 4).
Biomechanical Analysis of Fixation Methods
Two cadaveric studies have established that triangular osteosynthesis has superior stability compared with other methods of fixation for the treatment of vertically unstable sacral fractures37,38. Schildhauer et al. compared twelve cadaveric lumbopelvic specimens with a unilateral transforaminal sacral fracture and ipsilateral pubic rami fractures37. Six specimens were stabilized anatomically with an iliosacral screw, and six specimens were treated with triangular osteosynthesis, an iliosacral screw with lumbopelvic fixation from the L5 pedicle to the ilium. Schildhauer et al. reported a significant decrease (p = 0.0104) in initial displacement under peak loads for the specimens treated with triangular synthesis (0.163 cm) compared with the specimens stabilized anatomically with an iliosacral screw (0.611 cm). Additionally, after undergoing 10,000 cycles of loading, all specimens with triangular osteosynthesis had minimal motion, whereas for the specimens with iliosacral screw fixation, three (50%) had catastrophic failure and two (33%) had gross motion37.
In another study using a synthetic pelvic model, Berber et al. compared the stability of stabilizing a vertical fracture pattern with an iliosacral screw, a tension-band plate, an iliosacral screw and a tension-band plate, or a modified triangular osteosynthesis construct. Iliosacral screws were the weakest construct, failing after a load of only 50 N; comparatively, tension-band plating could withstand a load of up to 100 N before failing. Combining the two increased the strength, but it was still weaker than triangular osteosynthesis, which did not fail until a load of 500 N was applied38.
Clinical Use of Spinopelvic Fixation in Sacral Fractures
Evaluation and Diagnosis
Evaluation of patients with vertically unstable and complex bilateral sacral fractures should begin with a thorough Advanced Trauma Life Support evaluation, as these injuries are often the result of a high-energy mechanism. This should include both an anteroposterior pelvic radiograph and a computed tomographic (CT) scan of the pelvis, and it is critical to obtain coronal and sagittal reconstructions to evaluate for a transverse fracture and a kyphotic deformity. Additionally, the presence of concomitant injuries such as a closed head injury or a chest injury may delay or completely prevent the patient from undergoing a long surgery in the prone position.
Although there is no high-level evidence establishing which fracture types are best treated with spinopelvic fixation, multiple studies in the literature have demonstrated excellent radiographic outcomes in both unilateral and bilateral vertically unstable sacral fractures associated with lumbopelvic dissociation treated with spinopelvic fixation3,11,14-17. Isler identified that a unilateral vertical sacral fracture may exit lateral to the L5-S1 facet, through the facet or medial to the facet35, and fractures that exit through or medial to the facet lead to unilateral lumbopelvic dissociation. Spinopelvic fixation is our preferred technique in unilateral fractures with extensive fragmentation or fractures involving the L5-S1 articulation. Additionally, complex bilateral sacral fractures, such as U-type or H-type fractures, clearly benefit from spinopelvic fixation3,13,18.
Vertically Unstable Sacral Fractures
Vertically unstable sacral fractures present a treatment challenge because with previous treatment options, it was difficult to apply compression without increasing the risk of iatrogenic nerve root injury17. Although, to our knowledge, there have been no large prospective studies comparing fixation methods for vertically unstable sacral fractures, multiple case series have been published demonstrating high union rates without secondary displacement from spinopelvic fixation for these fractures3,11,14-17. Comparatively, the rate of secondary displacement is >20% when iliosacral screws are used in isolation for the treatment of complex sacral fractures39-41.
Sagi et al.17 reported on fifty-eight patients with acute hemipelvic instability from a comminuted, nonimpacted Denis Zone-II fracture with cephalad migration. All patients underwent unilateral triangular osteosynthesis, and no formal sacral laminectomy was performed. Partial weight-bearing was allowed immediately, and full weight-bearing was permitted at six weeks. Sagi et al. reported no signs of loss of reduction, and thirty-seven (93%) of forty patients with a one-year duration of follow-up demonstrated radiographic union on CT evaluation17.
Schildhauer et al. reported on thirty-four patients who had a vertically unstable sacral fracture and underwent spinopelvic fixation. All patients underwent triangular osteosynthesis and spinopelvic fixation with either tension-band plating or sacroiliac screws, and no patients underwent a formal sacral decompression. Patients were able to advance weight-bearing as tolerated. The earliest reported full weight-bearing was at eight days postoperatively, and the average time to full weight-bearing for patients without another contraindication was only twenty-three days. Two patients required a revision secondary to a loss of reduction, but all fractures united4.
Complex Bilateral Sacral Fractures
Complex bilateral sacral fractures are rare injuries, occurring in only 3% (thirteen of 442) of pelvic fractures18. Because of this, there is a paucity of literature available to guide physicians’ treatment algorithm. Currently, the largest series is by Schildhauer et al., who reported on eighteen patients with a Zone-III type-2 to 4 fracture treated with spinopelvic fixation3. Of these eighteen patients, isolated spinopelvic fixation consisting of bilateral L4 and L5 pedicle screws with two bilateral iliac screws was performed on eleven patients (61%), and triangular osteosynthesis consisting of bilateral L4 and L5 pedicle screws with one iliac screw bilaterally and one sacroiliac screw bilaterally was performed on seven patients (39%). All patients had severe neurologic deficits with complete loss of bowel and bladder function, and consequently, all patients underwent a formal sacral decompression. Furthermore, all patients also underwent a posterolateral fusion to enhance long-term stability. Postoperatively, all patients were permitted to bear weight as tolerated, unless precluded by other injuries. The average kyphosis decreased from 43° preoperatively to 20° postoperatively, and no loss of reduction or nonunion was reported3.
König et al. performed a systematic review of the literature and found twelve small case series reporting a total of sixty-three cases of U-type sacral fractures13. Of the sixty-three cases, twenty-three (37%) were managed with spinopelvic instrumentation and nine (14%) were managed with triangular osteosynthesis. No comparative analysis between fixation methods was performed, but they did report that thirty-five (97%) of thirty-six patients who had undergone an open surgery had fracture-healing without loss of reduction13.
Role of Decompression
The utility of a formal decompression with a sacral laminectomy is unclear. Although preoperative nerve deficits are common with vertically unstable and complex bilateral sacral fractures, there have been multiple studies that have shown patients with partial or even full recovery without a decompression7,11,12,18. Furthermore, to our knowledge, there have been no studies designed to examine the role of decompression. Although the available studies are difficult to compare because of heterogeneous techniques used for decompression (some authors perform a limited decompression through the fracture site11, and other authors perform a full sacral laminectomy3), and inconsistent reporting on the time to decompression, there does appear to be both a prognostic benefit and possibly a neurologic benefit of decompression.
In one study by Shildhauer et al., twenty-two (65%) of thirty-four patients with vertically unstable sacral fractures had a neurologic deficit on presentation; no patients underwent a formal decompression, and, of these twenty-two patients, only six (27%) had any signs of neurologic recovery4. Comparatively, in another study by Schildhauer et al., eighteen patients with a U-type sacral fracture presented with complete bowel and bladder dysfunction; after a sacral laminectomy and spinopelvic fixation were performed, fifteen patients (83%) had at least partial neurologic recovery3. Furthermore, the decompression can aid in determining the patients’ prognosis, as Schildhauer et al. reported that six (86%) of seven patients with all sacral roots intact regained complete bowel and bladder function, compared with four (36%) of eleven patients with one or more sacral root transected3.
Postoperative complications are common when treating complex pelvic fractures. In a systematic review of the literature, König et al. reported an overall complication rate of almost 40% for the treatment of bilateral sacral fractures with a transverse component13. Painful implants are the most common complication, with Sagi et al. describing fifty-five (95%) of fifty-eight patients reporting painful implants within one year; all symptoms resolved with removal of these devices17. The prominence of the iliac bolts can be decreased by countersinking the screws or using an anatomic starting point3,42,43. The rate of implant loosening is instrumentation-dependent. Up to 18% of iliosacral screws may loosen in triangular osteosynthesis3,17, whereas the rate of spinopelvic instrumentation loosening is <5%4,17. Importantly, to our knowledge, no cases of catastrophic failure of spinopelvic fixation for the treatment of sacral fractures have been reported. Wound complications and infections are other common complications, with the overall risk being as high as 26%1,3,17; the risk is higher in patients with degloving injuries1,3,17. Infections and wound breakdown in these cases are serious problems that necessitate a formal irrigation and debridement; however, with aggressive surgical wound care, infection eradication and primary wound-healing are possible1,3,17. Because of the rarity of these injuries, the exact rate of less common complications, such as iatrogenic nerve injury, is difficult to determine; however, in both studies with more than thirty patients, iatrogenic nerve injuries were reported (3% to 13% of patients)3,17. Lastly, Sagi et al. reported that six (15%) of forty patients had persistent coronal tilting of the L5 vertebral body on S1 even after implants had been removed17; however, to our knowledge, no long-term studies have been performed to determine the clinical ramifications of this.
Authors’ Preferred Operative Technique
The authors prefer open anatomic reduction of the sacrum with neural decompression, either through the fracture site or through a sacral laminectomy, and triangular osteosynthesis. In patients with bilateral spinopelvic dissociation, such as U-type or H-type sacral fractures, bilateral L4 and L5 pedicle screws with two iliac bolts and a transsacral screw are used (Fig. 5)3; in patients with a unilateral vertically unstable pelvic fracture that will have limited contact after compressive fixation (i.e., highly comminuted fractures) and all fractures through or medial to the L5-S1 facet, unilateral triangular osteosynthesis with an L5 pedicle screw, a single iliac bolt, and a transsacral screw is sufficient (Fig. 6)17.
Patients with a static neurologic examination are taken to the operating room in an urgent, but not emergent, manner after resuscitation with a lactate level of <2.5 mmol/L44. Patients who have progressive neurologic deterioration should be taken emergently to the operating room. Patients are placed prone on a specially designed pelvic reconstruction orthopaedic fracture table. A perineal post is used in complex bilateral fractures, but in patients with a unilateral vertically unstable pelvis, no perineal post is placed. Instead, a bicortical 5-mm Schanz pin is placed in the intertrochanteric region of the femur on the unaffected side and is attached via an external fixator to a pelvic skeletal stabilization frame. This rigidly stabilizes the patient without interfering with imaging or reduction45. Chest rolls and a transverse shoulder roll are used, allowing the abdomen to hang freely.
Most often, a midline incision is made from approximately L4 to S4 and is carried down to the lumbodorsal fascia. Next, a midline fascial incision is made, and subperiosteal dissection is performed from L4 to the distal sacrum, extending lateral to the transverse processes and the posterior superior iliac spines. Alternatively, in simple unilateral fracture patterns, either a long paraspinal incision can be placed ipsilateral to the displaced sacral fracture, and dissection between the multifidus and longissimus can be used for access to the posterolateral elements of the sacrum and lumbar spine; or by releasing the fascial origin of the gluteus maximus from the posterosuperior iliac spine and the dorsal fascia of the erector spinae, the erector spinae can be elevated laterally to medially, exposing the dorsal sacrum and sacroiliac joint46. Additionally, although not the senior author’s (M.D.S.) preferred option, Williams and Quinnan recently reported that the use of percutaneous lumbopelvic fixation may be safe and efficacious47. Appropriate fracture exposure with periosteal preservation is paramount. Any devitalized muscle is sharply debrided. Special attention is used during dissection around the facet capsules, as damage to these structures may increase the risk of degenerative changes and may lead to possible auto-fusion.
Once the exposure is completed, the wound is copiously irrigated to remove the fracture hematoma, and the fracture lines are carefully delineated. Next, a formal sacral laminectomy from S2 to S5 is performed. A curet and a pituitary rongeur can be used to isolate and to gently remove dorsal fracture fragments. Next, a Penfield 4 (Life Instruments, Braintree, Massachusetts) is used to protect the sacral nerve roots, and a Kerrison rongeur (Life Instruments) is used to complete the sacral laminectomy. Commonly, at this level, the nerve roots are no longer surrounded by the thecal sac, so no dural repair is needed. The sacral roots are then traced forward and any compressive bone fragments are removed. Depending on fracture morphology, neural identification and decompression may also be performed through the fracture site. Any traumatically severed nerve roots are sharply debrided. Throughout the rest of the case, protection of the L5 and intact sacral nerve roots is critical, as iatrogenic neurologic injuries can occur during fracture reduction. Similarly, at the conclusion of the case, care must be taken to perform a meticulous closure, as lumbodorsal fascia is the primary protection of the nerve roots.
Once the nerve roots have been decompressed and have been protected, the fracture can be reduced. In the setting of a unilateral vertically unstable pelvic fracture, a 5-mm Schanz pin is placed in the posterior superior iliac spine of the unaffected side and is connected to the two skeletal stabilization frames (Fig. 7). This rigidly stabilizes the hemipelvis and allows for traction to be applied through the injured side. (This step is not used in patients with U-type sacral fractures, as a perineal post is in place, and traction must be applied through both lower extremities. The fracture is clamped and compressed.) A combination of longitudinal traction through the fracture table and extension of the hips allows for a preliminary reduction of the hemipelvis. Intraoperative fluoroscopy is used to verify that the hemipelvis is no longer translated superiorly. Next, a small incision is made just lateral to the posterior superior iliac spine bilaterally, and a periarticular reduction clamp is placed. Prior to compression, the nerve roots are again inspected to ensure that they will not be entrapped in the fracture, and then the posterior sacrum is compressed through the clamp.
Once the reduction has been verified on fluoroscopy, a small stab incision is made to allow for percutaneous placement of a transsacral screw. With use of anteroposterior, inlet, and outlet pelvic radiographs, a 3.2-mm oscillating drill is used in a pistoning motion to drill transversely across the sacrum and perpendicular to the fracture line. The path should be cephalad to the S1 foramen on the pelvic outlet radiograph and must not violate that anterior sacral cortex on the pelvic inlet view to prevent iatrogenic injury to the L5 nerve root. Once the drill has been placed into the contralateral ilium, the drill bit is exchanged for a guidewire. A partially threaded cannulated screw with a washer is then placed that further compresses the sacral fracture. During compression of the fracture, it is critical that the nerve roots are well visualized and are protected from becoming incarcerated in the fracture.
Following insertion of the transsacral screw, the spinopelvic instrumentation can be placed. In a U-type sacral fracture, bilateral pedicle screws are placed at L4 and L5 with two iliac bolts (Fig. 5, D)3; however, in a vertically unstable pelvic fracture, often a unilateral L5 pedicle screw (if there is a fracture through the L5 pedicle, an L4 pedicle screw is used) and a single iliac bolt can be placed (Fig. 6, D)17. The starting point for the L5 pedicle screw is at the midpoint of the transverse process just lateral to the pars interarticularis. The L5 pedicles are often medially angulated 25°; however, preoperative templating is performed, which allows for the surgeon to know the exact size, length, and angulation of the pedicle. A high-speed burr is used to establish a starting point, followed by a pedicle probe to cannulate the pedicle. A ball-tip probe is used to verify that there is no cortical breach; finally, the pedicle is under-tapped by 1 mm, and a pedicle screw is inserted. Anteroposterior and lateral fluoroscopic imaging is used to verify the position of the pedicle screw. Next, with use of the 3.5-mm oscillating drill and starting at the internal aspect of the posterior superior iliac spine, a path for the iliac bolts is drilled. The path should aim approximately 20° lateral and 30° caudal43, so that the screw is between the tables of the ilium, just above the greater sciatic notch. A ball-tip probe is used to verify that there has been no cortical breach and the path is tapped. A 7.0-mm screw long enough to extend past the greater sciatic greater notch is placed, and the position is verified on anteroposterior, lateral, and rollover obturator oblique fluoroscopic images. A rod is then contoured and is inserted to connect the lumbar pedicle screws to the iliac bolts. If there is facet involvement or comminution, then posterolateral arthrodesis and bone-grafting are also performed.
In some complex fracture patterns in which there are multiple horizontal fractures, there can still be anterior displacement of the caudal-most portion of the sacrum after the upper sacrum has been reduced. This can be reduced with use of shoulder hooks and can be held in place with pointed reduction clamps. Conventional 3.5-mm pelvic reconstruction plates can then be placed lateral to the foramen on the distal aspect of the sacrum (Fig. 6, D). The wound is closed in layers with monofilament absorbable interrupted sutures, and the skin is closed with monofilament non-absorbable vertical mattress sutures. The fascia is closed over 2 g of vancomycin powder48, and a superficial and deep surgical drain is placed.
Postoperatively, the patient is allowed to bear weight as tolerated while using a walker for protection. After six weeks, the patient can slowly advance weight-bearing without the walker. In patients with devitalized tissue, a planned repeat irrigation and debridement is performed forty-eight to seventy-two hours after the index procedure. Because most often no fusion is performed, patients will undergo removal of spinopelvic fixation after radiographic union is verified on a CT scan. If a fusion is performed, the hardware is only taken out if it is symptomatic.
Vertically unstable and complex bilateral sacral fractures are devastating injuries with a high complication rate from both the injury and the treatment. Although the available evidence makes it difficult to develop a strict treatment algorithm, triangular osteosynthesis provides the strongest construct and allows for early weight-bearing with little concern for secondary displacement. Furthermore, a formal sacral laminectomy has prognostic value and may aid in neurologic recovery.
Source of Funding: There was no source of external funding for this study.
Investigation performed at the Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, Chicago, Illinois
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. In addition, one or more of the authors has a patent or patents, planned, pending, or issued, that is broadly relevant to the work. Also, one or more of the authors has had another relationship, or has engaged in another activity, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2015 by The Journal of Bone and Joint Surgery, Incorporated