Preliminary evaluation of a novel method for computed tomography quantification of lumbosacral articular process displacement in dogs with and without degenerative lumbosacral stenosis

Abstract

ObjectiveThis study aimed to describe the diagnostic discrimination and reliability of a novel technique for quantifying lumbosacral articular process displacement (LSAPD) on dorsal plane computed tomography (DPCT) imaging in dogs with and without degenerative lumbosacral stenosis (DLSS).Study designDPCT surveys of the lumbosacral vertebral column were performed with dogs positioned in extension and flexion. LSAPD is defined as the distance between the cranial aspects of the L7 and S1 articular processes. The LSAPD ratio is identified by dividing the LSAPD by the length of the L7 articular process. Intraclass correlation coefficients (ICCs) for intra- and inter-observer reliability were calculated, and logistic regressions were used to test for the association of LSAPD and LSAPD ratio with odds of DLSS. Significance was set at 0.05. Receiver operator characteristic (ROC) curves were calculated to determine diagnostic discrimination and optimal cutoff for LSAPD and LSAPD ratio in the diagnosis of DLSS.ResultsIntra- and inter-observer reliabilities were excellent for most measurements. In the current cohort, excluding covariates, the area under the curve (AUC) (95%CI) for LSAPD and LSAPD ratio measured in a flexed position were both 0.89 (0.82–0.96), suggesting potentially excellent discrimination for using this measurement as a marker for diagnosing DLSS, pending further studies. The cutoffs for flexed LSAPD and LSAPD ratio that maximizes Youden’s index were ≥ 1.2 mm and ≥ 9%, respectively. When age and weight were subsequently included as covariates in a multivariable analysis, a significant relationship between LSAPD or LSAPD ratio and odds of diagnosis of DLSS was not demonstrated, suggesting the need for a larger sample size.ConclusionThe results of this study suggest that measurements of LSAPD and LSAPD ratio on DPCT are feasible and reliable, although their diagnostic discrimination in DLSS should be evaluated further in future prospective studies.