Navigated percutaneous pelvic sacroiliac screw fixation: Experimental comparison of accuracy between fluoroscopy and Iso-C^3D navigation

Abstract

Percutaneous sacroiliac screw fixation is technically demanding and can result in complications mainly related to imaging problems. Furthermore, the conventional technique performed using fluoroscopic control is associated with a long radiation exposure. The purpose of this study was to evaluate the accuracy of two navigation technologies used in traumatology; fluoroscopy and Iso-C^3D navigation. A total of 40 screws were placed (20 with Iso-C^3D, 20 with 2D fluoroscopy) at levels S1 and S2. With both technologies, all S1 screws could be placed correctly, but four (10%) incorrect placements were seen at S2 with fluoroscopy navigation. With all Iso-C^3D navigated drillings, no perforation was seen. Iso-C^3D navigation therefore proved superior to 2D fluoroscopy navigation for sacroiliac screw fixation in an experimental set-up designed to assess accuracy.

• 3D navigation
• fluoroscopic navigation
• sacroiliac screw
• accuracy
• orthopaedic navigation
• screw placement

Introduction

In the treatment of unstable C-type pelvic injuries [1], osteosynthesis with iliosacral screws is a minimally invasive option for stabilizing the posterior pelvic ring [2–5]. Using the conventional fluoroscopy-based technique, the drilling can only be controlled in one projection at once; the position of the screws must be followed and adjusted under fluoroscopy in anterior-posterior (a.p.), inlet, outlet and lateral projections. This leads to increased radiation exposure for the surgeon and patient, and injuries to nerve roots and the gluteal vessels have been reported in 3–15% of cases using this method [6], [7].

More recently, navigation systems have been introduced for traumatological indications, especially in the field of spine and pelvic surgery [8–12]. Particularly in spine surgery, with CT-based navigation, the accuracy of pedicle screw placement could be improved [13], [14]. In unstable pelvic ring injuries, however, CT-based navigation is not applicable, as the fragment positions in the CT dataset may not reflect the intraoperative situation. For these applications, fluoroscopy- and Iso-C^3D-based navigation are indicated. The technology of these navigation modalities has been explained elsewhere [15], [16]. For both modalities, the images are acquired intraoperatively and represent the actual situation, e.g., after reduction of a fracture. However, each navigation technology has its advantages and pitfalls (Table I) [17].

Table I.  Comparison of fluoroscopy and Iso-C^3D navigation.

Navigation technology	Image quality	Actuality	3D	Accuracy
Fluoroscopy-based	+	+ + +	−	1–3 mm
Iso-C^3D	+ + +	+ + +	+ + +	< 1 mm

Basically, both modalities represent the actual situation in traumatology as compared to CT-based navigation. The accuracies are estimated and represent clinical experience.

The purpose of this study was to evaluate the accuracy of fluoroscopy and Iso-C^3D navigation when used for navigated sacroiliac screw fixation.

Methods

The navigation system SurgiGATE® (Praxim-Medivision, Grenoble, France) was used, including an Ultra 10 Workstation (Sun Microsystems, Palo Alto, CA) and an Optotrak 3020 optoelectronic localizer (Northern Digital Inc., Waterloo, Ontario, Canada) with an accuracy of 0.1 mm in translation and 0.1° in rotation, according to the manufacturer.

All drillings were performed by one examiner, an experienced surgeon, after extensive training in conventional and navigated sacroiliac screw drilling. Drilling was performed with a navigated drilling machine (Powerdrive, Synthes, Bochum, Germany) and new drill bits of 4.5 mm diameter. A navigated drill sleeve was used. After drilling, a k-wire (2.8 mm, Synthes) was inserted into the drill hole and a canulated titanium screw (7.3 mm, Synthes) was placed.

A total of 40 drillings were performed using 10 foam pelvises (Synbone, Malans, Switzerland), Twenty drillings (10 S1, 10 S2) were performed with fluoroscopy-based navigation, and 20 (10 S1, 10 S2) with Iso-C^3D navigation. For the drilling, the pelvis was fixed in a special holding device in the supine position, the pelvis itself being covered such that the examiner was blinded.

Postoperative evaluation was done by direct inspection and evaluation of CT images (Figures 1 and 2). The evaluation scored the outcome as follows: 0 = no perforation; 1 = minimal perforation without clinical relevance; 2 = obvious perforation (see Figure 1). The postoperative data set was analysed in the DICOM viewer eFilm Workstation™ (Merge eFilm, Milwaukee, WI).

Figure 1. Postoperative scoring of sacroiliac screw placement. A scoring system was developed to evaluate screw placement by direct visualization and by CT. a) Score = 0: no perforation. b) Score = 1: minimal perforation. c) Score = 2: obvious perforation.

Figure 2. Postoperative evaluation with CT. All screws were analysed with direct visualization as outlined in Figure 1 and by CT with axial, coronal and sagittal reformation. The screws shown in this CT image were placed under Iso-C^3D navigation.

Set-up for fluoroscopy-based navigated drilling

The specific “C-arm” module, available as one of the SurgiGATE® modules during the period of our study, was used (Figure 3). Four projections (a.p., lateral, inlet and outlet) were acquired with the navigated C-arm (Exposcope 8000®, Ziehm, Nuremberg, Germany). The images were imported into the system and the trajectories defined according to the relative course of the S1 and S2 screws.

Figure 3. Fluoroscopy-based navigation. For 2D navigation, fluoroscopy images were acquired for simultaneous a.p., inlet, outlet and lateral projection.

Set-up for Iso-C^3D-based navigated drilling

The Iso-C^3D Siremobil® (Siemens, Erlangen, Germany) was used (Figure 4). For the scanning the pelvis was placed on a purpose-built carbon table in the supine position. To plan the drillings it is necessary to determine exactly the volume of interest – in our series the center of S1 and S2. With a rotation of 190° around the isocenter, 100 single images were generated and converted simultaneously to 3D cut views. This data set was transferred via an interface (Navi-Link®, Siemens) to the navigation system in the mode of Iso-C^3D navigation. After planning the trajectories within the multiplanar planes, the navigated drilling commenced.

Figure 4. Set-up for Iso-C^3D navigation. a) After placing the reference base on the iliac crest, an Iso-C^3D scan was performed (100 images). b) The trajectories were planned in the 3D mode and drilling was performed. [Color versions available online.]

Results

With regard to the distribution of S1 and S2, it was found that all S1 screws could be placed correctly with a score of 0, and all incorrect placements were seen at S2. Sixteen (80%) of the fluoroscopy-based drillings were correct. Minimal perforation of the cortex (score = 1) was seen with 3 (15%) screws (S1 = 1, S2 = 2), and a more severe perforation (score = 2) with 1 (5%) screw (at S2). No perforation was seen with the Iso-C^3D-navigated drillings.

Discussion

Percutaneous sacroiliac screw fixation is a frequently used minimally invasive surgical technique. For safe positioning of the screw, the drill bit must be controlled in four views [18]. As this surgical technique is associated with radiation exposure for the surgeon, screw malpositioning can cause nerve injuries.

Navigation systems were introduced to increase the accuracy of surgical procedures and to decrease radiation exposure during operations. It was previously shown that use of navigation systems could reduce radiation exposure in other procedures [19], and an experimental study involving implantation of a dynamic hip screw found that the radiation dose could be reduced by 93% [20] with navigation.

Our results show that a higher proportion of sacroiliac screws could be placed accurately with Iso-C^3D navigation than with conventional C-arm-based navigation. Of course, the 100% perfect placement of screws with Iso-C^3D navigation cannot be transferred from a laboratory set-up into the clinical situation, but the difference when compared to fluoroscopy-based navigation is clear.

As the experimental set-up and surgical devices were identical for both series, the most likely reason for the difference in accuracy is one image source (the C-arm) having lower basic accuracy. It is suggested that the four views specified by Matta [5] may even increase the error of fluoroscopy-based navigation. However, 2D navigation has been successfully used for percutaneous screw placement in the pelvis [21]. Another clear advantage of Iso-C^3D navigation is the availability of a 3D tool for precise, postoperative 3D analysis. However, the additional costs of a 3D C-arm and associated equipment such as a radiolucent table must be mentioned.

It should be emphasized that the results from this experimental set-up cannot be transferred directly into the clinical situation. Initial clinical applications reveal that the image quality with both techniques is impaired by obesity and bowel gases.

A clear limitation of Iso-C navigation is the limited data volume (119 mm³) and the problem of metallic artefacts if other implants are already present. In comparison to a CT of the pelvis, orientation is difficult, especially when trajectories are to be planned. When drilling, the images may flip according to the drill-bit axis. However, CT-based navigation is not a practical alternative due to the unknown actuality of the image data set when the fracture is manipulated. An elegant CT modality-based and registration-free navigation system has been described by Jacob et al. [22]. However, this set-up is cost-intensive and can only be realized interdisciplinarily.

Conclusion

Iso-C^3D navigation demonstrates superiority to fluoroscopy navigation for sacroiliac screw fixation in an experimental set-up designed to assess accuracy.