The first model for viral-mediated gene therapy for SGN degeneration was reported by Staecker et al. [28]. In this study, BDNF-expressing HSV-1 vector was infused into the scala tympani in mice after complete destruction of the auditory hair cells by exposure to an aminoglycoside. Inoculation with BDNF-expressing HSV-1 vector resulted in a significant improvement in neuronal survival of SGNs at 4 weeks after infusion. Stable expression of the exogenous BDNF was observed 4 weeks after injection, suggesting that HSV-1 vector-mediated expression of BDNF contributed to SGN survival. The following studies, published from 2000 to 2004, demonstrated that different combinations of a neurotrophic factor and a viral vector (GDNF-expressing Ad5, BDNF-expressing Ad5, and BDNF-expressing AAV2) also enhanced SGN survival in ototoxically deafened guinea pigs at 4 weeks after inoculation into the scala tympani, indicating that viral-mediated long-term expression of neurotrophic factors is effective in preventing cell death of SGNs following the loss of hair cells [29, 31, 32]. Transgene expression was observed mainly in SGN cell bodies, suggesting that an autocrine mechanism may account for increased SGN survival. Regarding the delivery route of the virus, Wise et al. demonstrated that in comparison with scala tympani inoculation, injection of virals into the scala media resulted in greater SGN survival in the basal turn of the cochlea, which was associated with consistent transgene expression within the partially degenerated organ of Corti. However, since fewer SGNs survived in the middle and apical turns following injection into the scala media when compared with scala tympani inoculation, further experiments are necessary to reach a definite conclusion regarding the best injection site. Paralleling these gene therapy studies, researchers were attempting to combine cochlear implantation with viral-mediated overexpression of a neurotrophic factor to maintain the number of SGNs and improve CI outcomes. Rejali et al. coated the CI electrode with allogeneic fibroblasts transfected by BDNF-expressing Ad5, instead of using a neurotrophin-eluting biopolymer. Insertion of the BDNF-expressing electrodes preserved significantly more SGNs in the basal turns of the cochlea in guinea pigs for at least 48 days following implantation when compared to control electrodes [33]. Since direct infusion of virus into the inner ear may induce aversive immune responses, this ex vivo transfection approach has the potential to accomplish long-term growth factor secretion with minimal side effects.

Previous studies reported that peptide neurotrophic factors infused into the scala tympani using a mini-osmotic pump promoted SGN neurite extension and enhanced SGN survival [21, 24, 25]. In 2010, focusing on the effects of gene therapy on neurite extension of SGNs, two groups demonstrated that viral-mediated overexpression of BDNF and/or NT-3 resulted in resprouting of peripheral SGN fibers [35, 36]. It should be noted that regrowth of SGN fibers induced by an osmotic pump-mediated administration of neurotrophin peptides was disorganized, and looping back within the osseous spiral lamina and lateral projection along the basilar membrane were often observed [25]. The peptide infusion into the perilymph led a high concentration of administrated neurotrophins throughout the cochlea, which may have disturbed the original local gradient of neurotrophins released from the organ of Corti and caused the disorganized regrowth of SGN fibers. Consistent with this conclusion, inoculation of the neurotrophin-expressing virus into the scala media resulted in more localized transgene expression within the organ of Corti compared with scala tympani infusion and was associated with sprouting of SGN fibers to neurotrophin-expressing cells [35, 36]. The organized regrowth of SGN fibers induced by gene therapy clearly contrasts with the disorganized growth induced by osmotic pump-mediated administration of neurotrophin peptides, suggesting the advantage of viral-mediated gene therapy through scala media inoculation to promote regrowth of SGN fibers.

In the aforementioned studies, aminoglycoside-treated guinea pigs or rats were used to study gene therapy to prevent SGN degeneration secondary to ototoxic drug-induced hair cell loss. Recently, BDNF-expressing gene therapy was applied to Pou3F4 or conditional Gjb2 knockout mice, which are animal models for congenital genetic hearing loss [38, 39]. In particular, pups of Pou3F4 knockout mice have no cochlear hair cells [38]; therefore, these mice serve as a valuable model to evaluate whether viral-mediated BDNF expression can induce nerve fiber regeneration and SGN preservation in ears with hereditary deafness. Inoculation of BDNF-expressing AV into the scala tympani or scala media enhanced SGN survival in the basal turn of the cochlea in both strains of mutant mice. Regenerative sprouting of peripheral SGN fibers into the auditory epithelium was observed in the treated Pou3F4 knockout mice. These results suggest that congenitally deaf children with a hereditary genetic mutation can be candidates for gene therapy to improve outcomes of cochlear implantation.

In clinical application, the primary goal of gene therapy using neurotrophic factor-expressing viruses is to improve CI outcomes by preserving SGNs, which are the targets of CI-mediated electrical stimulation [20]. As described above, inoculation of BDNT, NT-3, and/or GDNF-expressing viruses into the scala media or scala tympani has been proven to prevent SGN degeneration and to enhance resprouting of SGN fibers in ototoxically and genetically deaf animals. However, it is important to determine whether these histological changes contribute to functional benefits in patients receiving CI-mediated electrical stimulation.

Chronic electrical stimulation in the cochlea alone significantly reduced deafness-related loss of SGNs when compared with unstimulated ears [40]. Electrical stimuli induced elevation of intracellular Ca²⁺ concentrations and release of synaptic vesicles containing neurotrophins and neurotransmitters from presynaptic terminals on SGNs via voltage-gated calcium channels, which may have prevented SGN degeneration [41, 42]. When the administration of BDNF peptide into the scala tympani using a mini-osmotic pump was combined with chronic electrical stimulation in ototoxically deafened guinea pigs, electrical stimulus thresholds were significantly lower than those measured in the ears receiving electrical stimulation alone, suggesting functional advantages of the infusion of BDNF peptide in cochlear implantation [43]. Consistent with this data, inoculation of GDNF- or BDNF-expressing virus into the scala tympani enhanced SGN survival and decreased EABR thresholds [30, 34]. The difference in the electrically evoked auditory brainstem response (EABR) thresholds was significant in the gene therapy experiment using BDNF-expressing Ad5 [34]. While continuous administration of neurotrophic peptides using a mini-osmotic pump carries the risks of infection, degradation of peptides in the pump, and cannula clogging, only a single injection is necessary to achieve long-term transgene expression in viral-mediated gene therapy, suggesting the advantage of gene therapy in clinical applications to improve CI outcomes. However, inoculation with viruses involves risks of an adverse immune response and toxicity, especially when using the virus the second time [44]. Accordingly, more extensive investigations are necessary before this technology is safe for clinical application in humans.