Between July 2010 and October 2010, 12 patients with vertically unstable sacral fractures were treated with pelvic external fixation and preand post-surgical longitudinal skeletal traction. Of the 12 patients, there were eight males and four females, aged range 19 to 52 years (average age 35.2 years). Of these patients, five cases suffered from traffic accident, four from fall injury and the other three from crush injury. Before the operation, the anteroposterior radiographs of the pelvis were performed. If patients were physiologically stable, CT scan was performed to identify the fracture types. These sacral fractures were all vertically unstable. According to Denis’ classification [8], two presented the zone I sacral fracture, nine cases the zone II sacral fracture and one case the zone III sacral fracture. At the initial examination, all cases were complicated by multiple fractures, including eight cases suffering disruption of the anterior pelvic ring (two cases with displacement). Five cases were complicated by traumatic shock, three cases by severe posterior soft tissue injuries. Neurological deficits were recorded before operation: two cases had L5 and S1 motor and sensory deficits, two cases had saddle anaesthesia, one case had bowel and bladder dysfunction.

After admission into our hospital, all patients were applied transtibial or transcondylar traction, with the initial traction weight of about 1/7 of their body weight. Immediate portable anteroposterior radiographs of the pelvis were reviewed to evaluate reduction accuracy at four hours interval. If the reduction was unsatisfied, 2 kg weight of traction was added until traction weight of about 1/4 of body weight. The operation was performed immediately, once the displaced fracture was reduced and the pre-operative preparation was completed (Figures 1(a) and (b)).

Under general anaesthesia, the patients were placed supine on a radio-transparent table with continuous longitudinal traction. The bilateral iliac crests were sterilely prepared. The first pin insertion site mounted on iliaccrest was approximately 2 cm behind the anterior superior iliac spine. A 1 cm incision was made to access the iliac crest. A drill guide was placed through the incision and down to bone. A 4.5 mm drill was placed through the guide and advanced approximately 10 mm, taking into account the oblique orientation of the iliac wing. Then a 6.0 mm pin was screwed at a depth of 4 - 5 cm. Other two pins were anchored behind the first one at the distance of 2 - 3 cm interval. Similarly, three more pins were placed in the contralateral iliac crest. The remaining components of external fixator were then installed on the pins. The fracture was then corrected by external manipulation under fluoroscopy (Figure 1(c)).

The patients with neurologic deficits and CT evidence of nerve entrapment by the bone flap needed the posterior surgical decompression and fixation, thus they were excluded in our study.

On the first postoperative day, pharmacological prophylaxis and intermittent pneumatic compression were applied to prevent venous thromboembolism (VTE) and pulmonary thromboembolism (PTE). The traction was applied for 4 - 6 weeks to maintain the position of the fracture. The external fixator was removed at the sixth to tenth week according to the type and severity of injury. They were permitted to walk with crutches without weight bearing or with partial weight-bearing. Three months later, they were encouraged to walk gradually in specialized rehabilitation department.

Functional outcome was evaluated according to scoring system proposed by Majeed [9], which was divided into seven factors such as pain (30 points), work (20 points), sitting (10 points), sexual intercourse (4 points), standing (36 points, including walking aid, gait unaided and walking distance). The total score then gave a clinical grade as excellent (≥85 points), good (70 - 84 points), fair (55 - 69 points) or poor (<55 points). The anteroposterior radiological examination of the pelvis was reviewed to evaluate the reduction of the fracture. The largest posterior displacement were recorded and graded according to the method proposed by Tornetta and Matta [10]. Displacement less than 5 mm was graded as excellent, 5 - 10 mm as good, 10 - 20 mm as fair, and more than 20 mm as poor. Statistical analyses were performed using SPSS for Windows 15.0.

External fixation is widely used in treatment of unstable anterior pelvic fractures in the initial stabilization of a hemodynamically unstable patient [15]. It also has a definite role to maintain reduction and finally stabilize pelvic injuries [16]. Safety and simplicity are the two major advantages in contrast to internal fixation [17]. However, there are few clinical reports and biomechanical analysis concerning treatment of sacral fracture by external fixation. Gunterberg et al. [18] carried out an experimental study on the loading of unilateral fractures of the sacrum and sacro-iliac dislocations with a trapezoid external compression fixation frame. The results indicated this frame properly reduced and stabilized these injuries. Meanwhile, it also gave sufficient stability to allow weightbearing in the upright standing position. Gardner et al. [19] confirmed that pelvic external fixation pins which were placed in the supraacetabular region imparted enough compression forces across the sacroiliac joint. Kim et al. [20] measured the stability of anterosuperior locations of pins into the iliac crest in pelvic injuries. They found that the external fixation could significantly reduce sacroiliac joint separation. Ilizarov external fixation device was also reported to efficiently treat unstable fracture of the posterior pelvic ring [21]. Based on these researches, it provides convincing evidence for the surgeons to stabilize vertical sacral fracture using anterior pelvic external fixation. In our consecutive series of 12 patients, the indirect reduction and stabilization were achieved by pelvic external fixation. At the final radiographical follow-up, ten patients were graded as excellent and good, and two fair. Our study confirmed that the application of external fixation was able to reliably maintain the stabilization of sacral fracture.

In the application of this technique, we have not used the technique of pins insertion into supraacetabular region in an anterior to posterior direction described by Gardner et al. [19], although it provides more compression [20]. Instead, we have performed pelvic external fixation perpendicularly anchoring these pins into each iliac crest in a superior to inferior direction, which has been used most frequently. Because former technique has potential disadvantages related to interference with the hip flexion, risk of hip joint penetration and damage to the lateral femoral cutaneous nerve branches [22,23].

It is important to restore the anatomic alignment of scrum fracture. However, it is a formidable task even by open reduction from posterior surgical approach [24]. Therefore, a controversial topic has arisen concerning whether anatomical reduction outcome is superior to functional reduction outcome [25]. Gansslen et al. [26] recommended anatomical reduction and decompression for the instable sacral fracture has instability. Many authors reported the indirect reduction could achieve the same excellent function result as the anatomical reduction [13, 27]. Chen et al. [11] reported excellence rate up to 85.7% according to the Majeed scoring system using indirect reduction and percutaneous internal fixation. In our study, complete or essential reduction of displaced fractures had been achieved. In the application of indirect reduction, a reasonable and large-weight longitudinal skeletal traction prior to operation should be administered to reduce the vertical displacement. The patients can endure traction weight of about 1/7 - 1/4 of their body weight, if the weight is added gradually. In our experience, we usually add 2 kg at four hours intervals. Repeated radiographic films are taken to make sure the extent of deformity correction. Postoperatively, the consistent traction is kept to maintain the reduction for average 4 - 6 weeks. Some authors advocated that pelvic external fixation combined with longitudinal traction badly affected the earlier weight-bearing and returned to their previous activity and work. However, it is not possible for patients with unstable sacral fractures, because they are frequently suffered from multiple fractures and neurological injuries [4]. With this method, all patients showed union of the sacral fracture and a satisfactorily stable pelvis. No secondary loss of reduction occurred. The patients had satisfactory functional results. Therefore, we made the primary conclusion that the iliac wing as a fixation site for lumbopelvic stabilizations in the pelvis and the preand post-operative continuous skeletal traction can restore and maintain the sacral fracture, and gain a good outcome to treat vertically unstable sacrum fracture.

As for neurological compromise, another controversial topic is whether surgical decompression of the neural elements improves better recovery results than conservative treatment [2,5]. Both methods have their supporters, but there is lack of solid evidence. Until recently, there have been no definite methods to accurately prove if the nerve roots are anatomical continuity. Therefore, many authors have suggested, only decompression of nerve roots is involved if debris or fragments compressed neural roots identified by CT scanning [4,11,28]. We support this viewpoint, so the sacral fractures in this case are excluded in our study and submitted to posterior operation. In our study, five of our patients were diagnosed with neurological deficit before operation. Only one did not completely recovered.

The major advantage of this method to treat fracture is in accordance to BO biomechanical conception. Potential benefits include minimizing soft tissue damage, shortening operation time, reducing perioperative complications as well as gaining faster rehabilitation compared with posterior open operation. In our research, the mean operative time was 38 min and the blood loss was rare. The closed reduction minimally interferes with the posterior soft tissue, hence avoiding the risk of infection. The major complication is pin sites infection, which can be eliminated by careful nursing and cleansing of the bases of the pins with dilute hydrogen peroxide twice daily. All 12 patients in our study had no pins track infection. Another vigilant complication is the development of VTE and PTE. Niikura et al. [29] reported that 19 patients (41.3%) were diagnosed with VTE, and PTE in ten (21.7%) in pelvic and acetabular fractures. The incidences were significantly higher compared with trauma patients without pelvic and acetabular fractures. Fortunately, all were asymptomatic. Administration of routine thromboprophy-laxis is fully justified [30]. In our study, pharmacological prophylaxis and intermittent pneumatic compression were applied to prevent venous thromboembolism. There was no form of symptomatic VTE and PTE.

This approach has several limitations. Pelvic external fixation has no effect on the transverse sacral fractures, though they constitute only 3% to 5% of all sacral fractures [31]. It also has limited function on the recovery of neurological compromise caused by compression of bone fragment, which we mentioned previously. The use of anterior external fixation should be under the circumstances of intact or the restoration of intact anterior pelvic ring. In this case, the incision of anterior operative approach and external pin sites will be conflicted. Despite the existence of these drawbacks, the anterior pelvic external fixation with skeletal traction is still an ideal surgical approach to treat vertically unstable sacral fractures.

Because this type of injury is relatively rare, our study represents a small cohort of patients. We are unable to compare the method of external fixation with the various operation methods via posterior approach by prospective and/or randomized studies. We also cannot draw the firm conclusion that external fixation with skeletal traction superior to another treatment method. However, we should not neglect its advantages regarding safety, efficacy and low complication.