INTRODUCTION
Displaced lesions of the pelvic ring including either a vertical fracture of the sacrum, or a dislocation of one or both sacroiliac joints, often pose problems of reduction and fixation. [1-5]
These lesions are optimally stabilized with internal osteosynthesis, which is performed, as a rule, by open surgery under visual control. [2, 4-7]
This type of procedure is rarely possible under emergent conditions. Joël Matta and Chip Routt [8] proposed a technique of percutaneous screw fixation. The former advocated the prone position, while the latter preferred the supine position. The advantages of this technique have been well described in the literature with, in particular, a net decrease in blood loss and risk of infection [8, 9] . The intermediate and long-term outcome of these fractures is held to be directly correlated to the quality of the reduction and to the absence of difference in lower limb length. Consequently, reduction of these fractures is the key element [10] . The problem still persists of certain reductions difficult to obtain with closed procedures, limiting the advantages of this technique [1, 2, 11] . Conscious of the advantages of percutaneous osteosynthesis, we propose a new reduction technique, limiting the necessity of an open approach to the fracture. This technique permits anatomical results in certain very displaced fractures, expanding the range of indications for percutaneous screw fixation.
SURGICAL TECHNIQUE
The operation is performed in three stages:
* The reduction is obtained by applying traction in the axis of the femur while blocking the trunk and contralateral limb. Reduction is indispensable before proceeding to the following stage.
* The entry point is determined either on lateral views as proposed by Routt, or in inlet and outlet views as proposed by Matta.
* Hardware placement is guided using AP inlet and outlet views.
- Patient positioning and reduction (Figure 1a)
Figure 1a: Traction on a conventional table. We have adapted a system of traction, which permits straightforward reduction and, above all, maintains it during the entire surgical procedure. The patient is placed in the supine position. Transcondylar traction is applied. The body of the patient is immobilized using several restraints. Only the thigh to which traction is to be applied should be unrestrained. The surgeon should then vary the degree of flexion of the thigh and knee to obtain optimum reduction.
The table base should not hinder displacements of the fluoroscopy C-arm necessary for the outlet views. On certain tables, the patient has to be moved toward the caudal end of the table to achieve adequate films. In large patients, the patient's feet may even extend beyond the table. Transcondylar traction is strongly recommended, because it relaxes the various nerves of the lower limb. In our opinion, hip flexion would also appear to contribute to reduction. The opposite foot is blocked by a support in such a manner that traction in the axis of the lower limbs acts to reduce the fracture of the pelvis.
One may also use an orthopedic table if it is radiolucent. However, on most orthopedic tables, the base blocks outlet views. On the Tasserit table, we add the leg supports to the main square without a countertraction post, placing the buttocks of the patient on these two supports. If the patient is not obese, this poses no problem and patient positioning is much simpler, making outlet views possible. Traction should always be transcondylar. The opposite foot is maintained in the foot support (figure 1b).
Figure 1b: The system of traction can also be set up on an orthopedic table provided that one can pass the C-arm for outlet views. In this image, the patient's buttocks are placed on two supports to leave enough room for the passage of the C-arm. This technique is possible only if the patient is neither too heavy nor too tall.
The quality of the reduction can be assessed on inlet and outlet AP views. Once reduction has been obtained, the pelvic region is rendered accessible and draped.
- Radiological location
The intervention takes place under fluoroscopic guidance.
The entry point of the iliosacral screw can be determined on lateral views, but this is only possible if anatomic reduction has first been achieved. If the reduction is imperfect (persisting displacement greater than 1 cm), open reduction is preferable, if possible.
To obtain adequate lateral X-rays, it is imperative to align two, bilateral anatomical landmarks from each iliac wing (Figure 2).
Figure 2: Lateral film of the sacrum. Lines a1 and a2 correspond to the radiological projection of the sacral ala, and lines b1 and b2 correspond to that of the greater sciatic notches. To obtain a true lateral view of the sacrum it is necessary for line a1 to overlap line a2 and for line b1 to overlap line b2. One may then define the area (in yellow here) where the iliosacral screw should pass. All of this is true only in a reduced pelvis. If the reduction has not been achieved, it is better to turn to an open technique. In the present series, all the patients operated within 48 hours had an excellent reduction. The small inserted image shows the relationships between the L5 nerve root and the bony region that defines line a1 in figure 2 on lateral films of the sacrum.
On a proper lateral X-ray of the sacrum, the two greater sciatic notches should overlap. This landmark was proposed by Routts. In our opinion, it is also helpful to align the radiological projection of the right and left sides of the pelvic inlet.
This second landmark on lateral views is interesting, because not only can it be used to ensure the quality of lateral views, but it also shows the position of the L5 nerve root in the sacrum. The L5 root is situated immediately anterior to this line on lateral views (Figure 2).
The entry point must be always situated below the projection of the pelvic inlet on lateral films.
A fine 10-cm pin is inserted in the outer table of the iliac wing. The image intensifier generally hinders direct placement of the definitive drill bit. Once the image intensifier in position face, one can replace the small pin by the definitive drill bit.
With experience, we remarked that lateral views were not essential and the entry point could be determined on the AP inlet and outlet views. These criteria were well defined by Joël Matta. During the intervention, it may be useful to secondarily verify the proper position of the screws on lateral views.