Lightning Podiums: Smorgasboard - Room 802B

Presented by: M. Lendner

Author(s):

A. Vaccaro(1), J. Harris(2), M. Hussain(2), R. Wadhwa(3), V. Chang(4), S. Schroerlucke(5), W. Samora(6), P. Passias(7), R. Patel(8), S. D'Agostino(9), N. Whitney((1)0), N. Crawford(2), B. Bucklen(2), M. Lendner((1)(1))

(1) Rothman Institute, Thomas Jefferson University Hospital, Department of Orthopaedic Surgery, Philadelphia, PA, United States
(2) Musculoskeletal Education of Research Center, Biomechanics, Audubon, PA, United States
(3) University of California, San Francisco Medical Center, Department of Neurological Surgery, San Francisco, CA, United States
(4) Henry Ford West Bloomfield Hospital, Department of Neurosurgery, West Bloomfield Township, MI, United States
(5) Tabor Orthopedics, Division of MSK Group PC, Memphis, TN, United States
(6) Nationwide Children's Hospital, Department of Orthopedic Surgery, Ann Arbor, MI, United States
(7) Hospital for Joint Diseases, NYU Langone Medical Center, Department of Orthopaedic Surgery, New York, NY, United States
(8) University of Michigan, Department of Orthopaedic Surgery, Ann Arbor, MI, United States
(9) Charleston Brain and Spine, Charleston, SC, United States
((1) 0) Inland Neurosurgery and Spine Associates, Spokane, WA, United States
((1) (1) ) Technion American Medical School (2) 0(2) 0, Spine Research Fellow | Rothman Institute, Philadelphia, PA, United States

Abstract

Introduction: Adult spinal deformity necessitates long posterior reconstruction to restore alignment. Anomalies related to deformity may affect pedicle screw accuracy. Navigation techniques aim to improve accuracy and reduce radiation exposure to clinicians; however, the efficacy of robotic navigation has not been thoroughly reported. The present study aims to quantify setup time, screw insertion time, intraoperative fluoroscopy, and accuracy of a novel robotic system in comparison to traditional open techniques in a cadaveric lab setting.

Methods: Ten surgeons (five neurological, five orthopedic) inserted eight pedicle screws from T10-L2 (robotic assistance on patient left, and conventional open on patient right; n=40 total per technique group), following pedicle exposure. The novel robotic system used included a camera system with markers for spatial tracking and a semi-active, rigid robotic arm. Intraoperative computed tomography scans were acquired and uploaded into the planning software; surgeons were not present and did not experience radiation during the initial scan. Recorded time for both techniques included exposure, setup, incisions, drilling, probing, tapping, and screw insertion. Accuracy was directly assessed via coronal, axial, and sagittal fluoroscopy following gross dissection and visual observation of each vertebral body. Pedicle breaches over 2mm were graded (Figure 1).

Results: Conventional open screw placement required an average 24.7±7.0 minutes; robotic assistance required 41.4±8.8 minutes per four screws (Table 1) (p= 0.485). Robotic navigation required 18.1 minutes of robot setup, intraoperative CT, and instrument verification. No differences were found between screw insertion times of conventional and robotic navigation (3.1±1.0 and 3.3±1.4 minutes, respectively; p= 0.982). Clinicians were exposed to 24.1±25.8 intraoperative fluoroscopic images necessary for traditional open screw placement; robotic navigation required no imaging following the initial scan (p= 0.012). Lastly, the conventional open technique resulted in a 42.5% (17/40) breach rate; robotic navigation resulted in no breaches greater than 2mm (p= 0.000). Robotic assistance observed a significant increase in screw length in comparison to the traditional open technique (50.3±4.1 and 44.0±13.8 millimeters); no differences were observed between screw diameters.

Conclusion: The use of a novel robotic-guidance system in a cadaveric setting significantly improved pedicle screw accuracy and eliminated surgeon radiation exposure. Setup of the robot system extended procedural time, but in comparison to the traditional open technique, this was not found to be significant.

Figure 1

Table 1 - Open Technique Comparisons