Comparison of Longitudinal and Transverse Approaches for Ultrasound-Guided Lumbar Erector Spinae Plane Block in Feline Cadavers

The ultrasound-guided erector spinae plane (ESP) block is a locoregional anesthesia technique primarily aimed at providing analgesia to structures innervated by the dorsal branches of the spinal nerves (DBSN). While this block has been widely studied in dogs, evidence in cats is limited, and only a few cadaveric studies have addressed the lumbar region. The aim of this study was to compare the injectate distribution and staining of the DBSN following ultrasound-guided lumbar ESP blocks performed using either a longitudinal or transverse approach in feline cadavers. A total of 15 feline cadavers were included, with 3 used for anatomical dissection and 12 for ultrasound-guided injections (24 sides). Injections were performed at the level of the third lumbar vertebra (L3) using a mixture of methylene blue, lidocaine, and iopromide (0.4 mL kg−1 per side). Needle placement and injectate spread were guided and confirmed by ultrasonography, followed by computed tomography (CT) and anatomical dissection to evaluate contrast and dye distribution. CT images revealed longitudinal spread of the contrast in all injections, with the transverse approach producing exclusively longitudinal distribution and the longitudinal approach showing occasional ventral spread beneath the transverse processes (25% of sides). Dissections demonstrated staining of a median of 2 DBSN per side for both approaches, predominantly L2–L3, with rare partial staining of ventral branches (VBSN) observed only with the longitudinal approach. The sympathetic trunk was not stained in any injection. No statistically significant differences were observed between approaches, except for L1 DBSN staining, which was identified significantly more frequently with the transverse approach (p = 0.033). These findings indicate that ultrasound-guided lumbar ESP block is anatomically feasible in feline cadavers and that both longitudinal and transverse approaches can result in injectate distribution to the DBSN.