Qualitative and quantitative computed tomographic characteristics of the lumbosacral spine in German shepherd military working dogs with versus without lumbosacral pain

Lumbosacral (LS) stenosis is a cause of lower back pain, loss of mission readiness, and early retirement in military working dogs (MWDs). Aims of the present two-part study were to evaluate a sample of German shepherd MWDs using standard clinical criteria for LS pain, standard qualitative computed tomographic (CT) criteria for LS stenosis, novel quantitative CT criteria for LS stenosis, and novel behavioral classification criteria for LS pain. Data were retrieved from archives of a tertiary referral MWD hospital. Study 1 was a retrospective, observational, two-group design with a hypothesis that there would be a significant difference in the percentage of affected German shepherd MWDs with multilevel stenosis (affecting ≥ 3 vertebrae) between LS pain groups, based on standard clinical and qualitative CT criteria. Study 2 was a retrospective, observational, cross-sectional, two- and three-group study design with a hypothesis that quantitative CT measurements would significantly differ between LS pain groups, assigned based on 3 classification systems. The 1st classification system used standard clinical criteria, while the 2nd and 3rd novel classifications included behavioral signs of LS pain. The following quantitative CT measures were recorded without knowledge of behavioral classification: vertebral foramen area, vertebral foramen volume, vertebral foramen fat area; and ratios of vertebral foramen area/vertebral body area (foramen area ratio), cranial vertebral foramen area/caudal vertebral foramen area (cranial:caudal foramen area ratio), and vertebral fat area/vertebral body area (fat area ratio). Study 1 findings did not support the hypothesis in that there was no significant difference in the percentage of dogs affected with multilevel stenosis between LS pain groups (P = 0.6567). Findings for study 2 supported the hypothesis in that dogs with LS pain were significantly more affected by multilevel stenosis (P = 0.0273). Significant differences occurred between LS pain groups in select vertebral locations for all measurements (P ≤ 0.05) except vertebral foramen area and vertebral foramen volume (P > 0.05). Comparisons using novel quantitative CT measures and behavioral classification criteria identified significant differences between LS pain groups that were not detected using standard qualitative criteria. These novel quantitative and behavioral classification criteria may be helpful in future research on causes for early retirement in German shephe