Feasibility and accuracy of intraosseous endoscopy for inspection of pedicle drill tracts in the canine thoracolumbar spine: an ex-vivo study
Mullins RA*1, Ortega C1, Bleedorn J*2, Maurin MP*1, Hoey S3, Espinel Ruperez J*4, Kraus KH*5, Hetzel S6, Guevar J7
1Department of Small Animal Surgery, Section of Small Animal Clinical Studies, University College Dublin, Dublin, Ireland, Dublin, Ireland, 2Colorado State University, Department of Clinical Sciences, Fort Collins, Colorado, USA, Fort Collins, USA, 3Equine Clinical Studies, Diagnostic Imaging And Anaesthesia, School of Veterinary Medicine, University College Dublin, Dublin, Ireland, Dublin, Ireland, 4School of Veterinary and Life Sciences, College of Veterinary Medicine, Murdoch University, Murdoch, Western Australia, Australia, Murdoch, Australia, 5Department of Clinical Sciences, College of Veterinary Medicine, Iowa State University, Ames, Iowa, USA, Ames, USA, 6Department of Biostatistics & Medical Informatics, University of Wisconsin‐madison School of Medicine & Public Health, Madison, Wisconsin, USA, Madison, USA, 7Vetsuisse Faculty, University of Bern, Division of Surgery, Bern, Switzerland, Bern, Switzerland.
Objectives:
Objectives were to evaluate feasibility of endoscopic inspection of pedicle drill tracts in the canine thoracolumbar spine and its accuracy for detection of breached / non-breached tracts.
Methods:
CTs of two greyhound cadavers from T6 to sacrum were obtained. Fifty-six pedicles (T7-L7) were randomised to have drill tracts with 1 of 5 modified Zdichavsky grades (non-breached, partial/full medial breach, partial/full lateral breach). Tracts were created using 3D-printed guides. Intraosseous endoscopy was performed using 1.9-mm 0-degree needle arthroscope. Grading was performed separately by 2 board-certified surgeons in a randomised blinded fashion. Postoperative CT grading of tracts was performed by a board-certified neurologist, blinded to endoscopy results. Specificity (ability to correctly detect a non-breached tract), sensitively (ability to correctly detect a breached tract), positive (PPV) and negative (NPV) predictive values were compared between surgeons.
Results:
Postoperative CT confirmed 43 non-breached tracts, 6 medial breaches (partial/full), and 6 lateral breaches (partial/full). One tract was excluded because of guide misplacement. Intraosseous endoscopy was successfully performed in all 55 tracts. Sensitivity to detect a medial and lateral breach was 83.3% and 50.0%. NPV was 93.1%. Specificity was 94.2%. PPV for detection of a medial and lateral breach was 83.3% and 54.5%. No significant differences in these comparisons were identified between surgeons. Median (range) time taken to assign a grade was 118 (30-486) seconds.
Conclusions:
Endoscopic inspection of canine thoracolumbar pedicle drill tracts is technically feasible and associated with high specificity, high sensitivity/PPV for detection of medial breaches, but poor sensitivity/PPV for detection of lateral breaches.