Navigated versus conventional transfixation of AC joint injuries: Feasibility and accuracy

Abstract

Objective: Transfixation of the acromioclavicular (AC) joint is a well-established technique for treating Rockwood IV to VI lesions. However, several complications, including pin breakage or pin migration due to incorrect placement, have been reported in the literature. A cadaveric study was performed to investigate whether the use of 3D navigation might improve the accuracy of AC joint transfixation.

Methods: Seventeen transfixations of the AC joint (8 non-navigated, 9 navigated) were performed minimally invasively in cadaveric shoulders. For the navigated procedures, a 3D C-arm (Ziehm Vision FD Vario 3D) and a navigation system (BrainLab VectorVision) were used. Reference markers were attached to the spina scapulae, then a 3D scan was performed and the data transferred to the navigation system. Two Kirschner wires (K-wires) were placed either freehand under fluoroscopic control (in the non-navigated group) or with the use of a navigated drill guide. Radiological analysis was performed with OsiriX software, measuring the distance of the K-wires from the center of the AC joint. For statistical analysis, Student's t-test was performed, with the significance level being set to p < 0.05.

Results: The maximum distance of the K-wires from the center of the AC joint was 5.4 ± 1.1 mm for the freehand non-navigated group and 3.1 ± 1.6 mm for the navigated group (p = 0.0054). The minimum distance of the K-wires from the AC joint center was 3.0 ± 0.6 mm for the freehand group and 1.6 ± 0.6 mm for the navigated group (p = 0.0002). The radiation time was significant lower for the freehand group (41.25 ± 20.4 seconds versus 79.5 ± 13.3 seconds for the navigated group, p = 0.004). There was no statistical difference between the groups with respect to the time required for surgery (11.25 ± 3.6 min for the freehand group and 12.6 ± 4.6 min for the navigated group; p = 0.475). In the freehand group, the AC joint was penetrated by both K-wires in 87.5% of the procedures, compared to 100% in the navigated group. Both K-wires were placed completely intraosseously in the clavicula in 50% of the procedures in the freehand group, compared to 88% in the navigated group.

Conclusion: Three-dimensional navigation may improve the accuracy of AC joint transfixation techniques. However, the radiation time is increased when using the navigated procedure, while the overall operation time remains comparable. Nevertheless, a 3D C-arm with a variable isocentric design is recommended for the acquisition of the shoulder scans.

• Acromioclavicular joint
• transfixation
• C-arm
• fluoroscopy

Introduction

Injuries of the acromioclavicular (AC) joint represent approximately 4% of all joint dislocations [1], [2]. Such injuries were classified by Tossy et al. in 1963 [3], with the classification being updated by Rockwood and Young in 1991 [4].

Until now, there have been no consistent treatment concepts reported in the literature for AC joint injuries [2]. There is a consensus that conservative treatment is appropriate for Rockwood I and II injuries [5–9], but the appropriate treatment for Rockwood III injuries remains controversial [5], [6], [10]. While in the United States most Rockwood III injuries are treated conservatively [11–15], in Germany the majority of patients with such injuries are subjected to surgery [5], [6]. Nevertheless, many authors agree that Rockwood IV–VI injuries should be treated operatively [5], [6].

A survey by Bäthis et al. [5] of the preferred treatment options for AC joint injuries in Germany showed that 39% of such injuries were treated by AC joint transfixation, 22% with the use of hook plates [16], and 9% with Bosworth screws [17]. Another survey by Powers and Bach [18] also found that 60% of the surgeons preferred transarticular stabilization of the AC joint as the main treatment option.

The most common complications after surgery are wound infections [19], [20], failure of the implant [19], and iatrogeneous fractures of the clavicula or acromion [21]. Particularly for transfixations of the AC joint, life-threatening dislocations and breakage of the implant have also been described [22–24].

Over the past few years, three-dimensional (3D) fluoroscopy has been shown to improve the precision of surgical tool placement in a variety of surgical procedures, often in combination with computer navigation [25–35]. Several studies have shown that the accuracy of procedures may be improved, and operative time, costs and the need for prolonged set-up reduced compared to conventional techniques or those using CT-based navigation [25], [32], [36–39]. Additionally, CT-based navigation offers no information concerning the patient's position on the operating table.

Nevertheless, 3D scans of the shoulder region using C-arms with an isocentric design have been reported to be difficult, if not impossible, to perform due to the isocentric design [31], [40]. Therefore, a flat panel radiation detector (FD) 3D C-arm has been developed (Ziehm Vision FD Vario 3D, Ziehm Imaging Gmbh, Nuremberg, Germany) in order to overcome these known limitations. In comparison to conventional C-arm systems (for example, the Iso-C^3D from Siemens Medical, Erlangen, Germany), the FD C-arm uses a variable isocentric design and the detector panel is much smaller. Image-quality output is 1024 × 1024 pixels for 2D and 512 × 512 pixels for the 3D model, compared to 256 × 256 pixels for the Siemens Iso-C^3D. The FD C-arm employs pulsed fluoroscopy with an output power of 2000 W; the X-ray tube voltage is variable from 40 to 110 kV; and the X-ray current intensity ranges from 0.2 to 20 mA [32].

So far, however, no information about the use of this technology for AC joint transfixations has been published. We hypothesized that 3D C-arm navigation would provide greater accuracy in K-wire positioning than the freehand technique, and conducted a human cadaveric study to evaluate the accuracy and feasibility of using a navigation system for minimally invasive procedures of this type.

Materials and methods

Five human cadavers (3 male, 2 female; median age 67 years) with no known deformities or previous injuries to the shoulder girdle were used for the present study. The cadaveric bodies were positioned in a beach-chair position on a radiolucent table. The shoulders were randomized for either the conventional technique with Kirschner-wire (K-wire) transfixation or the navigated procedure.

For the navigated procedure, the standard reference array (BrainLab VectorVision, Heimstetten, Germany) was positioned on the spina scapulae with two 3.0-mm Schanz screws pointed caudally, so as not to interfere with the K-wires for the transfixation (Figure 1 shows the set-up for the navigated procedure). An optoelectronic camera was then positioned so as to be able to track both the C-arm and the reference array [32]. A 3D C-arm (Ziehm Vision FD Vario 3D, Ziehm Imaging Gmbh, Nuremberg, Germany) was used initially to obtain 2D images to verify the correct positioning of the 3D C-arm.

Figure 1. Laboratory set-up. (a) Positioning on the radiolucent table with the flat detector 3D C-arm. The Kirschner wire is navigated via a navigated drill guide. The reference marker is attached to the spina scapulae. (b) Screenshot of a navigated AC joint transfixation. The upper left panel shows the aiming trajectory; the other panels show the planned trajectory.

Once the set-up and registration process were completed, image acquisition was initiated by performing a 3D scan. The radiation source of the C-arm then produced pulsed fluoroscopy while the C-arm was moved automatically through a scan angle of 136°. The 3D reconstructed images were created on the screen and then transferred to the navigation system for further intervention and navigation with the use of a navigated drill guide (BrainLab VectorVision, Heimstetten, Germany) (see Figure 1).

For the freehand technique, the procedure was performed with the aid of intraoperative 2D fluoroscopy. In this procedure, two 2.0-mm K-wires should be placed centrally in the AC joint without perforation of the outer cortex of the clavicle [2]. For the navigated technique, the K-wires were inserted through a navigated drill guide.

After implantation of the K-wires, another 3D scan was performed for documentation purposes and to permit further analysis. The radiologic analysis was performed using the OsiriX DICOM shareware viewer. Evidence of K-wire misplacement outside the AC joint was recorded, and the distance of the wire from the center of the AC joint determined, as well as any penetration of the cortex of the clavicle (see Figure 2).

Figure 2. Radiological analysis with the use of the OsiriX software. Multiplanar reformations are created, and the distance of the K-wires from the center of the AC joint is measured.

Statistical analysis

Student's t-tests were performed using GraphPad Prism (GraphPad Software, San Diego, CA). The significance level was set to 0.05.

Results

A total of 17 transfixation procedures (8 conventional, 9 navigated) were performed by an experienced orthopaedic surgeon (T.S.).

The maximum distance of the K-wire from the center of the AC joint was 5.4 ± 1.1 mm for the freehand group and 3.1 ± 1.6 mm for the navigated group (p = 0.0054) (Figure 3). The minimum distance from the AC joint center was 3.0 ± 0.6 mm for the freehand group and 1.6 ± 0.6 mm for the navigated group (p = 0.0002) (Figure 4).

Figure 3. Maximum distance of the K-wires from the center of the AC joint. Statistical analysis revealed significantly lower values in the navigated group.

Figure 4. Minimum distance of the K-wires from the center of the AC joint. Statistical analysis revealed significantly lower values in the navigated group.

Total radiation time was significantly lower for the freehand group (41.25 ± 20.4 seconds versus 79.5 ± 13.3 seconds for the navigated group; p = 0.004) (Figure 5).

Figure 5. Radiation time in the non-navigated and navigated groups. For the freehand group the radiation time covers only the intraoperative fluoroscopic imaging, while the significantly higher radiation time for the navigated group represents 2D fluoroscopic imaging along with the 3D scan.

There was no statistical difference between the two groups with respect to the overall time required for surgery (11.25 ± 3.6 min for the freehand group and 12.6 ± 4.6 min for the navigated group (p = 0.475) (Figure 6). In the freehand group, both K-wires penetrated the AC joint in 87.5% of the procedures, compared to 100% in the navigated group. In the freehand group, both K-wires were placed completely intraosseously in the clavicula in 50% of the procedures, compared to 88% in the navigated group.

Figure 6. Surgery time in the non-navigated and navigated groups. The comparison revealed no statistical differences. The time for the navigated group included 60 seconds for the 3D scan and approximately 1 minute for the transfer of the 3D data from the 3D C-arm to the navigation system.

Discussion

In our cadaveric study, 3D C-arm navigation with a flat detector panel significantly improved the overall accuracy of AC joint transfixation compared to a freehand technique. The advantages of multiplanar visualization throughout transfixation, as displayed by the navigation system, made the procedure more manageable. Nevertheless, higher accuracy was achieved at the expense of longer radiation time due to the 3D scan, though the operation time remained comparable.

AC joint transfixation for Rockwood IV to VI injuries is an established technique [7], [41], [42] and may be performed using a minimally invasive approach. Operative treatment for Rockwood III injuries should be reserved for young adults with a high demand for shoulder function [43], [44]. Such a procedure depends on the accuracy of the implant placement to avoid complications such as infections [44], K-wire migration into the lung, heart or great vessels [23], [27], [34], [45–48], or breakage [49], [50] and secondary loss of reduction [10], [50].

Another possible complication is early acromioclavicular degeneration, which may not be caused by articular perforation of the K-wires as it has also been reported after conservative treatment and coracoclavicular fixation [43], [51].

Performing the transfixation procedure in a freehand manner may be easy, but the surgeon has to rely on tactile information and 2D fluoroscopic X-rays. In some cases, several attempts may be required to obtain satisfactory positioning of the K-wire; and in the clinical set-up, the clavicle must be maintained in the anatomical position by clamps or manual retention. Our results show that intraoperative 2D fluoroscopic visualization may not always be adequate, as both K-wires penetrated the AC joint in only 88% of the freehand group. Obtaining the lateral axillary view may be particularly demanding in the intraoperative set-up with a beach-chair table. The multiplanar reformations obtained by 3D scanning and presented by the navigation system in real time may be helpful for this procedure, as we noticed no failures to penetrate the AC joint in the navigated group. However, only 3D C-arm systems with a variable isocentric design can provide scans of the shoulder region [40].

K-wire penetrations of the clavicular shaft were detected in both groups, but the incidence was significantly lower in the navigated group. K-wire penetration may be explained by the limited field of view of 3D C-arm systems in comparison to computer tomographs [21], [40]. In the current model, the 3D data cube is also limited to 13 × 13 × 13 cm, thus providing 1 cm more detail in each plane than a standard 3D C-arm system, with the peripheral K-wire position being sometimes not visible due to the limited data volume [25].

As shown in our study, the accuracy of AC joint transfixation can be improved using 3D C-arm flat detector navigation. Accurate positioning of the K-wires, together with anatomical repositioning of the AC joint, is supposed to improve the biomechanical behavior of the repaired area. However, clinical data of this kind is not available in the literature and was not part of this study.

The present study had several limitations. First, the group size was rather small and inconsistent, with only 8 cadavers being used for the conventional approach and 9 for the navigated procedure. Second, the AC joint was not dissected prior to the procedure, leaving the ligaments intact for optimal accuracy analysis. Third, as already described above, the size of the data cube, which is provided by the flat detector 3D C-arm, is limited in comparison to a CT scan with variable data volume, and this may lead to misjudgment of implant position in a clinical setting.

Conclusion

Flat detector-based navigation seems to be promising for application in navigated AC joint transfixation. The accuracy of the procedure is improved, and while radiation time is increased, the overall time required for surgery remains comparable to that for the conventional technique. Misplacement of K-wires with respect to AC joint perforation occurred only in the freehand group, and not in the navigated group.

Declaration of interest: The authors report no declarations of interest.