Patient-specific drill template for C2 transoral pedicle insertion in complete reduction of atlantoaxial dislocation: cadaveric efficacy and accuracy assessments

The surgical treatment of atlantoaxial dislocation (AAD) presents a difficult surgical challenge for neurospinal surgeons [1, 2]. A transoral atlantoaxial reduction plate (TARP) was developed specially for AAD and became an effective advance in the treatment of AAD [3‐5] that could allow anterior atlantoaxial release, reduction, decompression, and internal fixation in a one-stage operation. Transoral C2 pedicle fixation was adopted in the third generation of the TARP system [6], which could effectively improve screw anti-pullout force and might offer higher stability than intravertebral insertion fixation. However, due to anatomical variation and the proximity to the vertebral artery and spinal cord, transoral C2 pedicle insertion (C2TOPI) is relatively difficult and has a high potential for injury of the arteries, nerve roots, and the dural sac [7].

Accurate C2TOPI is a key to successful clinical application of the TARP system. Because of the anatomical complexity of atlantoaxial structures and the area covered by the TARP, C2TOPI is a relatively difficult task. The anatomical freehand technique remains the mainstay in clinical practice. However, screw malposition occurs frequently [8‐10]. C-arm fluoroscopy was thus introduced to facilitate pedicle screw placement in upper cervical spinal surgeries [11, 12]. However, the overlapping images are not sufficient for atlantoaxial complexity and therefore cannot indicate the screw position precisely [13]. Li et al. [2] reported the placement of C2TOPI using the freehand technique under C-arm fluoroscopy, in which 46.9% of screw threads penetrated the bony cortex of pedicles. Further research revealed that the pedicle cortex penetration rate in unacceptable positions was as high as 13.0%; one patient died postoperatively as a result of C2 screw misplacement [1]. Furthermore, intraoperative fluoroscopy might extend the operation time and increase the total amount of radiation; this is dangerous for both the patient and the surgical team [14, 15].

Patient-specific drill templates (PDTs) produced by three-dimensional (3D) printing have been developed for spinal surgeries; this method could not only obviate radiation exposure and time-consuming procedures, but could also improve the accuracy in cervical posterior pedicle insertions [16‐20]. However, PDTs for C2TOPI are rare. To the best of our knowledge, only Li et al. [2] reported grouped PDTs in their cadaveric study, in which several graded screw trajectories were pre-set to facilitate C2TOPI. However, because the intraoperative reduction of AAD might not be perfectly fit for the pre-set graded screw trajectories, potential risks could be present in the clinical application of PDTs.

Intraoperative reduction of AAD can be divided into two situations, complete and incomplete reductions [4, 21], which require different strategies for PDT design. In this study, we developed a grouped PDT specifically for AAD with complete reduction and compared its efficacy and accuracy in facilitating C2TOPI. We hypothesized that this grouped PDT might be an alternative to the fluoroscopy-guided freehand technique for C2TOPI.

The PDTs were constructed successfully using 3D reconstruction and 3D printing. During the operation, the PDTs were fitted to their corresponding anterior cervical surfaces appropriately without any free movement. K-wires were easily inserted into the cervical pedicle with the assistance of the PDTs. The surgery time was 37.8 ± 4.5 min in the PDT group and 63.3 ± 3.82 min in the freehand group, resulting in a significant difference between the two groups (t = − 13.5, p < 0.001). The production time for this grouped PDT was approximately 70 min, and the manufacturing cost was approximately $30 per patient.

The absolute deviations from the centroids between the preoperative designs and postoperative measurements on the axial plane of the pedicle were 0.79 ± 0.64 mm in the PDT group and 1.69 ± 0.54 mm in the freehand group. On the sagittal plane of the pedicle, the corresponding values were 0.96 ± 0.51 mm in the PDT group and 1.72 ± 0.64 mm in the freehand group. Significant differences in the absolute deviations were observed on both the axial and sagittal planes (t = − 3.36, p = 0.003 and t = − 2.88, p = 0.01, respectively) (Table 1).

Eight (80%) screw positions were type I, and two (20%) were type II in the PDT group. In the freehand group, five (50%) pedicle screw positions were type I, three (30%) were type II, and two (20%) were type III (Fig. 5). The classification of the screw positions was not significantly different between the two groups (χ² = 2.553, p = 0.443) (Table 2).

In this study, we developed a grouped PDT specifically for C2TOPI in AAD with complete reduction; the efficacy and veracity were further validated by comparison to the fluoroscopy-guided freehand technique.

In this study, we mainly focused on C2TOPI for AAD with complete anatomical reduction. An ideal trajectory that is considered to not only takes into account screw insertion safety but also biomechanical properties could be preoperatively designed for this procedure. A PDT might be subsequently developed to facilitate C2TOPI. However, unlike the usual lock-and-key structures of PDTs using a massive bone surface [23‐25], this PDT was quite different. The TARP was not only an implant but also a reduction tool as well. After reduction of AAD, the TARP was temporally fixed with a VBS and covered most areas of the C2 anterior surface. Therefore, the PDT had two functions in facilitating C2TOPI. One was to facilitate an intraoperative reduction that allowed the TARP to find the ideal trajectory that was preoperatively designed. PDT A was designed for this purpose and allowed determination of the positions of the TRS, setscrew, and two C2TOPI entry points. In the intraoperative reduction, when the position of the bottom of the TARP reached the height of the TRS and the two entry points of C2TOPI were detected at the center of the pedicle screw holes of the TARP, the TARP was considered to have found the ideal trajectory and was then fixed with a VBS. The other purpose was to facilitate trajectory drilling that allowed the TARP to be fixed to C2 using transoral pedicle fixation. Because the TARP covered most of the area of the anterior surface of C2, the narrow uncovered surface, including the surface between two wings of the TARP and under the TARP, was used for location registration. PDT B, which included a table, the beneath registration columns/pins, and the above two guide tubes, was specially designed. PDT B was firmly attached to the anterior surface of C2 with good surface registration between the lower structures of columns/pins and the narrow surface not covered by the TARP. Trajectory drilling was then successfully performed using the guide tube. In the present study, the surgery time was approximately 1 h in the freehand group, which is slightly longer than most routine operations performed in the operation room. This result might have occurred because the surgeon was only an attending spinal surgeon with minimal experience in C2TOPI. He had to perform numerous fluoroscopic examinations to determine the ideal trajectory for C2TOPI; this required a lengthy time. However, for the same surgeon, the surgery time was much shorter in the PDT group. This finding indicates that the PDT could effectively improve the efficiency of young surgeons during their early attempts of C2TOPI. Further advantages of the PDT included reduced radiological exposure and the elimination of the need for complex equipment and complicated intraoperative procedures [16], especially for those with less C2TOPI experience.

The accuracy of PDTs in facilitating C2TOPI is of vital importance. In the present study, we calculated the absolute value of the deviations from the centroids between the preoperative designs and postoperative measurements, which had been used in our previous studies [25‐27] to show the actual deviations. Our study showed that the absolute deviations in the axial plane (0.79 ± 0.64 mm) and the sagittal plane (0.96 ± 0.51 mm) in the PDT group might be within an acceptable range for clinical application. Both of these results were significantly different from those in the freehand group.

The rank index of the screw position in C2TOPI is also very important and can be used to assess critical cortex penetration intuitively. A rank index of the screw insertion position was introduced to evaluate the outcome of C2TOPI [1, 2]: type II is considered an acceptable breach, and a type III perforation is considered an unacceptable breach. This classification might be reasonable. Anatomical research has shown that the tolerance distances for C2TOPI consist of a space of approximately 1.1 mm between the lateral pedicle wall and vertebral artery and a space more than 1.1 mm between the medial pedicle wall and dural sac [28, 29]. Further clinical research has confirmed that perforations less than 50% of the diameter of the screw pose no risk of neurovascular injury. In our study, although no significant difference was observed in the screw positions between the two groups, two unacceptable breaches (type III) occurred in the freehand group (20%). This rate was slightly higher than that in previous reports (13.0%) [1], most likely because the surgeon had limited experience in AAD treatment, and the sample size was relatively small. In general, the results including the intuitionistic rank and quantitative absolute values of the deviations revealed that the PDT-guided technique was more precise than the fluoroscopy-guided freehand technique in facilitating C2TOPI.

Although the specially designed PDT had obvious advantages in facilitating C2TOPI, some limitations still exist and may require further attention. First, we used the ideal trajectory in our study. This trajectory was only considered for morphological safety but was not considered to be biomechanically optimized. Thus, further research is needed. Second, the PDT involved in this study was designed only for facilitating complete reduction of AAD in C2TOPI. A different strategy is required for incomplete reduction of AAD because the trajectory cannot be determined preoperatively. We designed a different PDT for this procedure that will be reported at a later time. Finally, the PDT was only used on cadaveric spines. Clinical studies are needed.

In summary, specific PDTs could provide surgeons with an accurate and easy-to-apply method to facilitate C2TOPI for AAD with complete reduction. This could provide a viable alternative for surgeons.