Feasibility and accuracy of intraosseous endoscopy for inspection of thoracolumbar and lumbar pedicle drill tracts in a canine large-breed cadaveric model

Abstract OBJECTIVE To evaluate the feasibility of endoscopic inspection of thoracolumbar and lumbar pedicle tracts in a canine large-breed model and its accuracy for the detection of breached versus nonbreached tracts. ANIMALS 2 greyhound cadavers. METHODS CT scans of 2 greyhound cadavers from the sixth thoracic vertebra to the sacrum were obtained. Fifty-six pedicles were randomized to have drill tracts with different modified Zdichavsky grades (nonbreached, partial/full medial breach, or partial/full lateral breach) using 3-D–printed guides. Endoscopy was performed on a single occasion from October 9 to 10, 2023, using a 1.9-mm 0-degree needle arthroscope in a randomized blinded fashion. The grading of drill tracts was performed on postoperative CT. Specificity, sensitivity, positive and negative predictive values, and time to assign endoscopic grade were investigated. RESULTS Postoperative CT confirmed 43 nonbreached tracts, 7 medial breaches (partial/full), and 5 lateral breaches (partial/full). One tract was excluded because of guide misplacement. Intraosseous endoscopy was successfully performed in the remaining 55 drill tracts. Sensitivity to detect medial and lateral breaches was 71.4% and 60.0%. Negative predictive value was 93.1%. Specificity was 94.2%. Positive predictive value for detection of medial and lateral breaches was 83.3% and 54.5%. Median (range) time to assign an endoscopic grade was 118 (30 to 486) seconds. CLINICAL RELEVANCE Intraosseous endoscopy of pedicle drill tracts may be a useful adjunct technique during pedicle screw/pin placement in dogs.