Drug distribution along the cochlea is strongly enhanced by low-frequency round window micro vibrations

Abstract The cochlea’s inaccessibility and complex nature provide significant challenges to delivering drugs and other agents uniformly, safely and efficiently, along the entire cochlear spiral. Large drug concentration gradients are formed along the cochlea when drugs are administered to the middle ear. This undermines the major goal of attaining therapeutic drug concentration windows along the whole cochlea. Here, utilizing a well-known physiological effect of salicylate, we demonstrate a proof of concept in which drug distribution along the entire cochlea is enhanced by applying round window membrane low-frequency micro vibrations with a probe that only partially covers the round window. We provide evidence of enhanced drug influx into the cochlea and cochlear apical drug distribution without breaching cochlear boundaries. It is further suggested that ossicular functionality is not required for the effective drug distribution we report. The novel method presented here of local drug delivery to the cochlea could be implemented when ossicular functionality is absent or impeded and can be incorporated in clinically approved auditory protheses for patients who suffer with conductive, sensorineural or mixed hearing loss.


40
The relative inaccessibility of the human cochlea and its intricated structure requires new drug 41 delivery technologies to be designed to ensure safe, efficient and uniform drug distribution  48 and, to a limited extent, into the scala vestibuli through the oval window occluded by the stapes. 49 If the drug is allowed only to diffuse passively along the narrow, extended ST, its 50 concentration, in theory, should become the same within the entire scala after an arbitrary long 51 time (unrealistic scenario, Figure 1B) (Sadreev et al., 2019). However, for a drug to be 52 effective, it has to be cleared from the ST into other cochlear compartments (more realistic 53 scenario, Figure 1B). Dynamic equilibrium between diffusion and clearing leads to the 54 formation of substantial steady-state, base-to-apex drug concentration gradients along the   Figure 1. Schematic of the mammalian hearing organ (A) and two scenarios of molecular drug diffusion along the scala tympani (B). (B) Passive molecular diffusion of a drug along the scala tympani is described by a diffusion (kd) and clearing (kc) coefficients. For a given geometry of the scala tympani, the steady-state drug concentration gradient (denoted by the blue colour intensity) along it depends only on the ratio kd/kc (Sadreev et al., 2019). (A) is modified from (Lukashkin et al., 2020). 73 Animal preparation and signal generation and recording have been described elsewhere 74 (Burwood et al., 2017). Briefly, pigmented guinea pigs of similar weight (350-360 g) and both 75 sexes were anaesthetised with the neurolept anaesthetic technique (0.06 mg/kg body weight 76 atropine sulphate s.c., 30 mg/kg pentobarbitone i.p., 500 µl/kg Hypnorm i.m.). Additional 77 injections of Hypnorm were given every 40 minutes. Additional doses of pentobarbitone were 78 administered as needed to maintain a non-reflexive state. The heart rate was monitored with a 79 pair of skin electrodes placed on both sides of the thorax. The animals were tracheotomized 80 and artificially respired with a mixture of O2/CO2, and their core temperature was maintained 81 at 38°C with a heating blanket and a heated head holder.

85
Signal generation and recording 86 The middle ear cavity of the ear used for the measurements and salicylate application was 87 opened to reveal the RW. Compound action potentials (CAPs) of the auditory nerve in response 88 to pure tone stimulation were measured from the cochlear bony ridge in the proximity of the 89 RW membrane using Teflon-coated silver wire coupled to laboratory designed and built 90 extracellular amplifier (James Hartley). Thresholds of the N1 peak of the CAP at different 91 frequencies, which corresponds to different distances from the cochlear base (Greenwood,92 1990), were estimated visually using 10 ms pure tone stimuli at a repetition rate of 10 Hz.

93
For acoustic stimulation sound was delivered to the tympanic membrane by a closed acoustic 94 system comprising two Bruel and Kjaer 4134 ½" microphones for delivering tones and a single 95 Bruel and Kjaer 4133 ½" microphone for monitoring sound pressure at the tympanum. The 96 microphones were coupled to the ear canal via 1 cm long, 4 mm diameter tubes to a conical 97 speculum, the 1 mm diameter opening of which was placed about 1 mm from the tympanum.

98
The speculum was sealed in the ear canal. The closed sound system was calibrated in situ for 99 frequencies between 1 and 50 kHz. Known sound pressure levels were expressed in dB SPL re 100 2×10 -5 Pa.

101
All acoustic stimuli in this work were shaped with raised cosines of 0.5 ms duration at the    The probe vibration amplitude was calculated by integrating its velocity.

127
During experiments, the carbon probe was placed at about 45-degrees to the RW because of a 128 limited access to the RW. Probe vibrations started immediately after placing salicylate solution 129 on the RW. In the first 20 minutes period, acoustic CAP threshold recordings were taken every 130 3-5 minutes to record the fast action of salicylate at the basal region of the cochlea. Due to the 131 very low frequencies used to vibrate the RW, there was no CAP generated in response to the 132 probe vibrations, allowing recordings of the CAP due to acoustic stimulation to be taken during 133 the RW micro vibrations. After 20 minutes, the CAP threshold recordings were made every 10 134 minutes until a total of 60 minutes of RW micro vibrations. To washout, the carbon probe was 135 removed, the salicylate was removed from the RW using fine paper wicks and the recovery of 136 CAP threshold to acoustic stimulation was recorded. 138 Stapes vibrations were recorded using a laser vibrometer (CLV-2534, Polytec GmbH, 139 Waldbronn, Germany). The laser beam was focussed on the stapes head. The output voltage 140 from the vibrometer was low-pass filtered at 100 kHz, with a sensitivity of 2 mm/s/V.  The carbon probe touching the membrane was pushed slightly toward inside of the pipes at rest 159 to ensure membrane tension and its relaxation during backward phase of probe strokes.      Figure 5D). However, due to nonlinear dispersion of the diffusion front, this 267 small additional spread led to a statistically significant increase (unpaired t-test for 0.5 mm 268 bins, p<0.05) in the fluorescence intensity, i.e. dye concentration, over a much wider range of 269 9 mm (blue horizontal bar, Figure 5D). (1)

Recording of stapes vibrations
Even if a small far-field pressure was generated in our experiments due to finite stiffness, , of 301 the RW, which did not generate measurable stapes vibrations, then it still would not lead to a  experimental values is 6.82×10 -4 mm 2 /s), than we observed for Lucifer yellow (diffusion 347 coefficient is 3.1×10 -4 mm 2 /s (Brink & Ramanan, 1985)).

348
The second major difference between our in vivo and fluorescent dye experiments is in the 349 amount of material available for diffusion. The amount of dye was limited by its initial