Targeted inner ear delivery of gadolinium using microbubble-assisted ultrasound in an ovine model

Current strategies for treating sensorineural hearing loss include auditory rehabilitation, cochlear implantation, and systemic or local drug administration. Transtympanic injection (TTI) allows local drug delivery to the middle ear, but drug diffusion through the round window membrane (RWM) into the inner ear (IE) remains inconsistent. Microbubble-assisted ultrasound (MB-assisted US) has emerged as promising modality to enhanced RWM permeability. While feasibility and safety have been demonstrated in small animal models, translational validation in large mammals is necessary. This study aimed to compare gadolinium (Gd) diffusion into the IE following MB-assisted US versus passive diffusion in a sheep model, and to assess safety. Five normal-hearing ewes underwent bilateral mastoidectomy. One ear received Gd (Gadovist®) and Vevo MicroMarker® MBs (2.10⁷ MB/mL), followed by MB-assisted US exposure using a 1 MHz US probe (100-μs inter-pulse period, with 300-kPa peak negative pressure for 3-min exposure time). The contralateral ear received Gd via TTI. IE Gd diffusion was assessed by a serial MRI at 10, 20, 30 min and 7 days after Gd delivery. Auditory brainstem responses and vestibular function were evaluated at 1 h pre-treatment and at 7 days post-treatment; metabolomic analysis was performed on perilymph samples. Gd diffusion was greater with MB-assisted US than with TTI, with a 10- and 3.6-fold greater residual volume at 30 min and at Day 7 post-delivery, respectively. No auditory or vestibular toxicity was observed, and no metabolic alteration of the perilymph was detected. In conclusion, these findings support the translational potential of MB-assisted US for IE drug delivery.