A novel patient-specific three-dimensional-printed external template to guide iliosacral screw insertion: a retrospective study

Iliosacral (IS) screw fixation is a popular method for the management of posterior pelvic ring fractures or dislocations, providing adequate biomechanical stability [1]. The classic percutaneous technique, as described by Matta and Saucedo [2], uses fluoroscopic guidance to reduce the extent of soft tissue disruption, decrease the volume of blood loss and shorten the overall operative time. However, fluoroscopic guidance is a significant source of radiation exposure to both the patient and medical staff. Additionally, patient-specific factors, such as obesity, intestinal gas and a dysmorphic sacrum, have been associated with a screw malposition rate of 2–15%, which may lead to catastrophic neurovascular injury [3, 4].

The use of computer-assisted navigation systems has been advocated to lower the risk of screw malposition, and of associated neurovascular injury, as well as to decrease exposure to radiation during IS screw insertion [5‐7]. Although these systems do provide greater accuracy and less radiation exposure, their high cost and the complexity of the setup have limited the widespread application of these systems in intermediate- and primary-care hospitals.

Individualized internal templates, based on three-dimensional (3D) image reconstruction and reverse engineering, provide an alternative to computer-guidance for the placement of a K-wire to guide the accurate insertion of the IS screw (Fig. 1a). Although use of an internal template provides an accuracy in screw placement that is similar to that of computer-assisted navigation, it does require an incision of approximately 5 cm and soft tissue dissection to provide a clear bone surface to match the template [8, 9]. Moreover, in our experience, we have found that the geometry of the posterior iliac crest is inadequate to restrain the internal plate, leading to slippage and malposition of the IS screw. As well, soft tissue residues on the bone may also contribute to a mismatch between the template and bone surface, again resulting in malposition of the IS screw.

Emergent external fixation is widely used for the primary management of pelvic fractures. The external fixator pins, which are inserted deep into the iliac crest, could provide a simple geometric surface and solid cornerstone for mounting an external template. Therefore, we developed a low-profile 3-dimensional (3D) printed external template that can be firmly attached to the external fixator pins, overcoming the abovementioned disadvantages of internal templates, to guide IS screw insertion (Fig. 1b and Fig. 2). The aim of our study was to describe the use of our new patient-specific external template to guide the insertion of iliosacral screws and to evaluate the efficacy and safety of this technique compared to the conventional fluoroscopy-guided technique. We hypothesized that use of the 3D-printed external template would lead to a superior outcome, in terms of the accuracy of screw insertion, a shorter operative time and lower radiation exposure, to the conventional technique.

A total of 65 screws were implanted in 40 patients: 37 screws (19 S1, 18 S2) in the template group (22 patients) and 28 screws (14 S1, 12 S2) in the conventional group (18 patients). The quality of the reduction was not significantly different between the two groups (Table 2). The average operative time per screw was 17.9 ± 4.5 min (range, 17–52 min) for the template group, which was significantly less than the 39.7 ± 10.7 min (range, 63–115 min) required for the conventional group (p < 0.001). The radiation exposure dose was lower for the template (742.8 ± 230.6 cGy/cm²) than conventional (1904.0 ± 844.5 cGy/cm²) group (p < 0.001; Table 1). The rate of screw perforation was also lower in the template (1 of 37 screws, 1 at grade 1) than conventional (4 of 28 screws, 2 at grade 1, 2 at grade 2) group (P < 0.001). No incidence of neurovascular injury was identified among cases of screw perforation (Table 3). No incidence of pin site infection was noted among patients treated with emergent external fixation.

In the template group, the mean distance between the actual and virtual screws was 2.7 ± 0.8 mm (tip), 1.8 ± 0.6 mm (nerve root zone) and 1.5 ± 0.5 mm (entrance point), with a mean deviation angle of 1.8 ± 0.8°.

Use of our novel external template to guide K-wire insertion, and subsequent IS screw placement, provided high accuracy, with a shorter operative time and lower radiation exposure than the conventional fluoroscopy-guided technique. Moreover, despite the conventional use of inlet, outlet and lateral fluoroscopy views to guide percutaneous IS insertion, violation of the cortex remains an inherent problem of IS screw implantation [18‐20], with additional fluoroscopic views often required intra-operatively to further improve accuracy. Ozmeric et al. [19] advised that two different inlet views should be used to evaluate the anterior and posterior borders of the sacral body, separately. Kim et al. [18] even suggested using two inlet views (at 25° and 55° from the vertical) and three outlet views (at 25°, 35° and 55° from the vertical) to avoid misperception of the local anatomy. However, any adjustment in the position of the guide wire required in one view should necessarily be reconfirmed on all fluoroscopic views. As such, improving the accuracy of placement of the guide wires using the conventional method increases both intra-operative time and radiation exposure, which limits the use of this iterative approach for accurate guide wire placement in clinical practice. Different assistive tools have been used to improve the accuracy and efficiency in placing guide wires, including the use of multidimensional fluoroscopy [20], an internal template [8, 9], thermoplastic membrane navigation [21], and computer-assisted navigation [5, 6, 17, 22]. Our external template is distinctive from these assistive approaches, with greater pre-operative time taken to design and print the navigation template, which avoids the need for repeated fluoroscopy during the surgery and reduces the overall operative time.

Radiation exposure is a significant concern for percutaneous IS screw insertion. Excessive amounts of radiation can cause muscle weakness, hair loss, cataracts, and even cancer [23]. Computer-assisted navigation does allow a surgeon to exit the operative theatre during image capture, reducing radiation exposure for the surgeon (and medical team). A distinct advantage of our external template is the shorter intra-operative time required to insert the guide wire within the target corridor. As such, the dose of radiation exposure was significantly lower (at 742.8 ± 230.6 cGy/cm²), for both patients and the medical staff, compared to the conventional technique (at 1904.0 ± 844.5 cGy/cm²).

The accuracy of IS screw insertion has traditionally been evaluated using the penetration grades [15]. In our study, the penetration grade was significantly lower for the external template than conventional group (Table 3). However, due to the ceiling effect, this grading system cannot be used to assess the accuracy of the large number of screws that did not penetrating the bone cortex. Takao et al. [17] described a method to evaluate the deviation distance between the planned and actual IS screw, measured at the tip of the screw, the nerve root tunnel zone and entry point, on sagittal view CT images. Using 3D CT fluoroscopy matching navigation, they reported mean deviations of 2.2 ± 0.8 mm (tip), 1.8 ± 0.7 mm (nerve root tunnel) and 2.5 ± 1.8 mm (entry point). A similar method was adopted by Takeba et al. [22], using the O-arm navigation system, with a mean deviation in the position of the tip of the screw of 1.3 ± 0.6 mm. In our study, we used the Takao et al.’s method, with deviation measures comparable to those of the above studies, in which computer navigation was used to guide IS screw insertion. The satisfactory results we obtained with regard to deviation between the planned and actual screw placement and low rate of penetration into the cortex reflect the precise planning of the IS screw trajectory, which was strictly along the midline of the osseous corridor and was confirmed on axial, coronal and sagittal planes in software. Moreover, the special strategy wherein the template was designed such that it could be mounted on the contralateral hemipelvis helped avoid screw penetration in cases with poor fracture reduction. Furthermore, with experience, we have developed the following tips for designing the template to decrease the extent of deformity of template system. First, the template should have a low profile, which minimizes the distance between the template and the skin and, thus, between the marker pins and the virtual screws (Fig. 3d-e). Second, the distance between the two pins inserted in the iliac crest needs to be sufficiently wide to provide good anti-rotation ability for the external template. Third, the guide sleeve in the external template should closely match the guide wire and be sufficiently long to adequately guide K-wire insertion. Fourth, the template system should be constructed using a high-strength material, such as stainless steel or photosensitive resin, providing a strong restraint to bending.

The delay in definitive surgery because of the time required for the design and printing of the template is a limitation of our technique. The mean time from injury to surgery for the template group in our study was 5.4 days, compared to 4.1 days for the conventional group (p = 0.049). This longer delay for the template group, however, did not negatively influence the quality of the fracture reduction, which was comparable between the two groups. Of note, all definitive surgeries of both groups were performed within 14 days of the trauma, and the use of the pelvic reduction frame for closed reduction [13] partly offset any adverse effects of a longer delay to surgery.

Pin site infection, which is a common complication of external fixation [24, 25], is also a potential weakness of our external template technique due to the need to insert pin into the iliac crest. However, there was no incidence of pin site infection in our study group. Use of a strict cleaning protocol for the pin sites [26] and effort to shorten the duration of pin fixation [27] can be important factors in preventing pin site infection. Furthermore, the marker pin insertion is an extra process for patients who did not need emergent external fixation. The decision to use this technique needs careful consideration and informed consent from these patients. The indications for the use of our external template technique among patients who did not require emergent external fixation, include, but are not limited to, the need for insertion of multiple percutaneous screws, dysmorphic sacrum and osteoporotic pelvic fractures.

In summary, the external template is an effective tool for percutaneous IS screw insertion, decreasing radiation exposure for both surgeon and patients and avoiding neurovascular injury caused by screw malposition. Considering the high accuracy in IS screw placement that we achieved using an external template, we propose that this technique could potentially be used for other percutaneous insertions of screws in pelvic or acetabular surgery (Fig. 3e-f and Fig. 4c). Further applications and studies are still required to confirm this hypothesis.