OBJECTIVE High cervical spine fixation represents a challenge for spine surgeons due to
the complex anatomy and the risks of vascular and medullar injury. The
recent advances in 3-D printing have unfolded a whole new range of options
for these surgeons.
METHODS In the present study, a guide for the placement of the lateral mass screw in the C1 vertebra was developed using 3-D printing. Eight real-size models of the high cervical spine and their respective screw guides were built using computed tomography (CT) scan images. The guidewires were inserted with the help of the printed guides and then the models were analyzed with the help of CT scan images.
RESULTS; All of the guidewires in the present study obtained a safe placement in the models, avoiding the superior and inferior articular surfaces, the vertebral foramen, and the vertebral artery.
CONCLUSION The present study demonstrated the efficiency of the guide, a reliable tool for aiding the insertion of guidewires for screws in lateral masses of the C1.
Keywords: cervical vertebrae/pathology; cervical vertebrae/surgery; spinal fractures; spinal fusion; three-dimensional printing.
OBJETIVOS A fixação de coluna cervical alta pode representar um desafio para os
cirurgiões de coluna devido à anatomia complexa e aos riscos de lesão
vascular e medular. Os recentes avanços com a tecnologia de impressão 3 D
abriram um novo leque de opções para os cirurgiões.
MÉTODOS Desenvolveu-se umguia para a adaptação de parafusos demassa lateral em C1 comauxílio de impressão 3 D. Foram confeccionados oitomodelos em tamanho real de coluna cervical alta e seus respectivos guias com base em tomografias computadorizadas.Os fios-guia foram introduzidos com o auxílio dos guias; os modelos foram analisados com auxílio de tomografia computadorizada.
RESULTADOS Todos os fios-guia avaliados no estudo apresentaram um trajeto seguro nos modelos, respeitaram as superfícies articulares superiores e inferiores, o canal vertebral e a artéria vertebral.
CONCLUSÃO O estudo demonstrou que o guia tem boa eficácia, é uma ferramenta confiável para auxiliar a adaptação de fios-guia para parafusos em massas laterais de C1.
Palavras-chave: vértebras cervicais/patologia; vértebras cervicais/cirurgia; fraturas da coluna vertebral; fusão vertebral; impressão tridimensional.
|Citation: Nanni FN, Vialle EN, Foggiattob JA, Silva KWSNe, Mello Neto HO. Development of a Patient-speciﬁc Guide for High Cervical Spine Fixation*. 54(1):20. doi:10.1016/j.rbo.2017.09.011|
Work developed at the Hospital Universitário Cajuru, Curitiba, PR,
Felipe de Negreiros Nanni's ORCID is https://orcid.org/0000-0001-6959-8862.
|Received: Agosto 16 2017; Accepted: Setembro 05 2017|
High cervical spine ﬁxation always challenges spine surgeons. The complex anatomy, the importance of this region to cervical mobility,1 and the relatively small area for ﬁxation result in several technical difﬁculties.2 Initially, the ﬁxation methods were insufﬁcient, as the fusion techniques from Brooks-Jenkins or Gallie,3 and required prolonged external ﬁxation, or they were extensive, included the occipital region, and signiﬁcantly limited the mobility of the high cervical spine.
The introduction of the screw ﬁxation at the lateral masses, initially with Goel et al4 and then with Harms et al,5 changed the surgical spectrum for high cervical spine. However, the anatomical challenges persisted, mainly due to the location of the vertebral artery, which is lateral and superior to the entrance point of the screws,6 anterior to the carotid artery, medial to the spinal cord,7 and to the extensive venous plexus adjacent to the root of the C2 vertebra at the screw entrance point in the C1 vertebra.2 Moreover, this procedure requires continuous radioscopy, which is harmful both to the patient and to the medical team.8
The use of computed navigation emerged as an option to increase the accuracy of this method.9 However, its cost is elevated, and this technique is not available in most hospitals.
In the present study, the authors aimed to provide a low cost, easy-access option to reduce the risks associated with high cervical ﬁxation using a guide speciﬁc for the anatomy of the patient, developed with 3-D printing.
The objective was to evaluate the accuracy of the ﬁxation of the lateral masses of the C1 vertebra with a patientspeciﬁc guide made with a 3-D printer from a 3-D cervical model built from control computed tomography (CT) scans.
MATERIAL AND METHODS
The population of the present study consists of unidentiﬁed CT images from 8 adult patients > 18 years old. None of the eight patients presented extreme morphological alterations or signiﬁcant deformities at the high cervical spine.
High-resolution, thin-slice CT scans (01 mm) were selected. The ﬁlms were analyzed by the authors, who determined the absence of important deformities and the presence of preserved, signiﬁcant anatomical landmarks.
Preparation of the High Cervical Spine Models
Using the InVesalius software (CTI, Campinas, SP, Brazil), sequential two-dimensional CT images were converted in a 3-D model, which was then exported to CAD 3D Meshmixer (Autodesk Inc., San Rafael, CA, United States) and operated to isolate the C1 and C2 vertebrae, and a part of the C3 vertebra.
Cervical vertebrae models were printed in acrylonitrile butadiene styrene (ABS) with a 3-D Vantage I printer (Stratasys Inc., Eden Prairie, MN, United States) in a 1:1 scale (►Fig. 1).
Preparation of the Pedicle Fixation Guides
With the cervical spine models ready, a guide for pedicle screws was created. These guides were designed for stable adapting over the vertebrae, limiting the lateral or superior-inferior translation and trying to avoid anatomically important structures.
Using the CAD 3D Meshmixer software (Autodesk Inc., San Rafael, CA, USA), the virtual model of the guide was then assembled over each individual high cervical virtual reconstruction. A Boolean operation identiﬁed the intersection between the models and ﬁne adjustments were individually performed for the proper ﬁtting of the guides at the posterior surfaces of all of the evaluated cervical spines.
Each guide path was individually deﬁned with the help of a surgeon, simulating on the computer the desired track for each screw.
The guide models were then printed in acrylic resin Fullcure 720 with an Objet EDEN250 printer (Stratasys, Eden Prairie, MN, USA) (►Fig. 2).
Adaptation of the Guidewires
The printed guides were adapted over their respective printed cervical spine models. The cervical models ﬁtperfectly to the guides, which were stable at handling.
With the models ﬁrmly secured with a bench vise, and with the guides adapted and ﬁrmly positioned by the surgeon, both pedicles from the eight C1 vertebrae were perforated with a 1.5 mm drill. After the perforation, the drill was released from the equipment and remained ﬁxed to the model.
Analysis of the Results
After imaging the parts with the adapted guidewires, their position was measured by the RadiAnt DICOM Viewer software (Medixant, Poznan, Poland) based on four parameters: distance from the guidewire to the vertebral artery foramen; distance from the guidewire to the medullary canal; distance from the guidewire to the superior articular surface of the C1 vertebra; and distance from the guidewire to inferior articular surface of the C1 vertebra (►Fig. 5).
The CT scan images showed that none of the 16 guidewires from the 8 models invaded the medullary canal area or the articular surfaces, neither were at the region of the vertebral artery path.
The distance from the vertebral artery foramen ranged from 2.8 to 6.2 mm, with a mean value of 4.08 mm, a median value of 4 mm, and a standard deviation (SD) of 0.937 mm.
The distance from the vertebral canal ranged from 3.7 to 8 mm, with a mean value of 5.83 mm, a median value of 6 mm, and a SD of 1.266 mm.
The distance from the guidewire to the superior articular surface of the C1 vertebra ranged from 5.9 to 10.7 mm, with a mean value of 7.52 mm, a median value of 7.1 mm, and a SD of 1.212 mm.
The distance from the guidewire to the inferior articular surface of the C1 ranged from 2.0 to 5.9 mm, with a mean value of 3.62 mm, a median value of 3.5 mm, and a SD of 0.988 mm.
The values for each parameter are listed in ►Table 1.
|Model 01||Model 02||Model 03||Model 04||Model 05||Model 06||Model 07||Model 08|
Abbreviations: DAVD, distance to the vertebral artery to the right; DAVE, distance to the vertebral artery to the left;DCVD, distance to the vertebral canal to the right; DCVE, distance tothe vertebral canal to the left;DSAID, distance tothe inferior articular surface to the right; DSAIE, distance tothe inferior articular surface to the left; DSASD, distance to the superior articular surface to the right; DSASE, distance to the superior articular surface to the left.
Three-dimensional printing became a promising resource to help planning and executing complex spinal surgeries, allowing to recreate, with great accuracy, intricate anatomical models from imaging results.10-12
Recent studies investigate the efﬁcacy of 3-D printing in preparing patient-speciﬁc surgical guides to assist the adaptation of pedicle screws in lumbar and thoracic spine vertebrae. Moreover, there is evidence of good results with the intraoperative in vivo use of these guides.
In an experiment by Fu et al,16 polymethyl methacrylate guides for transpedicle screwsweremanuallymoldedover cervical vertebrae models prepared with a 3-D printer and tested in cadaveric vertebrae, with good results and easy applicability.
Some studies have also tried to develop pedicle screw guides for cervical vertebrae printed directly in 3-D, with excellent results. In an experiment by Sugawara et al,17 100% of 80 adapted screws in 20 patients were precisely located according to the preoperative planned path, with an average deviation of 0.29 ± 0.31 mm (0.0 mm-1.6 mm).
Studies regarding 3-D printing in vertebral spine surgery are still scarce and, until now, there is not a consistent study about these pedicle guides for the C1 vertebra in the literature.
The atlantoaxial region is a surgical challenge with unique anatomical and biomechanical properties. Several techniques were developed, but the ﬁxation of screws in the lateral mass of the C1 vertebra gained prominence during the last decade.18
There are technical difﬁculties to identify and access the optimal entrance point for the lateral mass screw in the C1 vertebra. For this reason, a careful preoperative planning is required to avoid the bad positioning of the screws and the excessive exposure that can cause expressive bleeding and harm the success of the surgery.18,19
Difﬁculties during the approach to the posterior aspect of the lateral mass of the C1 vertebra include abundant bleeding when exposing the inferior aspect of the posterior arch and the posterior portion of the lateral mass of the C1 vertebra,20 and the risk of vertebral artery lesion when mobilizing it in the groove of the arch of the C1 vertebra.6
The exposition of the nerve of the C2 vertebra and the careful dissection of the tissues adjacent to the dorsal ganglion of the root of the C2 vertebra should be performed.6
The screws are directly inserted in the lateral mass of the C1 vertebra, inferiorly to the base of the posterior arch. The dimensions of the lateral mass of the C1 vertebra easily accommodate 3.5 mm screws in most patients.19
According to a technique described by Harms et al,5 the screws must be inserted in a posterior-anterior direction, with 5º to 10º of convergence on the axial plane. At the sagittal plane, they must remain parallel to the caudal aspect of the posterior arch of the C1 vertebra, pointing to the center of the anterior tubercle of the C1 vertebra.6
During the studies to prepare these guides through computer-modeling, these anatomical aspects were carefully noted to deﬁne an individual, optimal path for each model.
The originality of the present study created some difﬁculties regarding the optimal format of the guide. Different guide models were prepared and tested to determine which one would present the highest stability when adapted to their respective vertebral models.
Analyzing the number ﬁgures for each vertebra, anatomical differences in morphology and size between the eight models were considered.
The guidewires did not invade the joint region, the medullary canal or the vertebral artery path of none of the eightmodels. All of the 16 guidewireswere passed only once, with no false trajectories.
All of the adapted guidewires satisfactorily kept their paths through the center of the lateral masses of the C1 vertebra and presented enough adjacent bone mass for a possible adaptation of a cannulated screw in a safe manner.
The guides were resistant to drilling, and there was no printed guide rupture or material deformity due to the heat generated by the burr. A small amount of debris originated from the guides was observed after drilling. Since the current literature lacks conclusive data about the long-term effect of this debris in living organisms, additional studies are warranted.
Due to the good results obtained with the guides, new studies will be performed to improve the guides and verify their possible in vivo applicability.
The experiment showed that patient-speciﬁc guides made with a 3-D printer allowed the positioning of guidewires in the lateral mass of the C1 vertebra with a precision of 100%. These guides are easily applied, can beneﬁt the adaptation of the screws in the lateral mass of the C1 vertebra, and provide a safe path for the screws.