Richard Salvi, E. Lobarinas, W. Sun, D. Stolzberg, G. Yang, L. Zhang
Center for Hearing & Deafness, University of Buffalo and Dept. of Communicative Disorders & Sciences, Buffalo, NY 14214, email: salvi@buffalo.edu

The goal of my presentation is to review our research on tinnitus with humans and animals over the past two decades and to discuss the interplay between the basic sciences and clinical research. Our initial electrophysiological studies on tinnitus using animals prompted us to carry out a series of brain imaging studies with humans. Our human brain imaging studies then led us back to studies with animals that were trained to reliably report on the presence or absence of tinnitus. In the past few years, we have used animal models to invetigate the mechanisms of
tinnitus at the cellular and molecular level and evaluate potentially useful drugs to treat tinnitus.

The phantom sound of tinnitus was traditionally believed to originate from damageinduced hyperactivity in the auditory nerve. However, the neural activity in damaged ears is generally unchanged or hypoactive in cases of severe damage to the inner hair cells. Despite a reduced neural output from damaged cochleas, the neural activity in the central auditory system is often hyperactive. This led to the hypothesis that the tinnitus generator(s) might reside in the central auditory pathway and possibly involve nonauditory regions.

To test this hypothesis, we used positron emission tomography (PET) to map the sites of brain activity in: (1) Subjects with somatic tinnitus that could modulate the loudness of their tinnitus with an oral facial movement (OFM). (2) Acoustic neuroma patients that could modulate the loudness and pitch of their tinnitus by eye movements (gaze-evoked tinnitus, GET). (3) Subjects whose tinnitus loudness increased or decreased in response to intravenous lidocaine treatment.

Using PET imaging, we showed that OFM produced a significant increase in activity only in the auditory cortex contralateral to the ear in which the tinnitus was perceived. In contrast, a real sound presented to normal subjects activated both the left and right auditory cortex. Since the pattern of neural activity from OFM-induced tinnitus was fundamentally different from that evoked by a real sound, we hypothesized
that the tinnitus generator was in the central auditory pathway. Similar conclusions were drawn from studies of GET and lidocaine studies. In GET patients, who experience tinnitus in their dead ear, lateral eye gaze evoked activity in the auditory brainstem or in regions adjacent to auditory cortex. In the auditory literature, lidocaine is touted as a drug that can temporarily suppress tinnitus; however, in the cardiology literature, tinnitus is listed as a potential sideeffect of intravenous lidocaine. In our study with normal and tinnitus subjects, intravenous lidocained produced bidirectional effects, in some subject tinnitus became quieter while in others it became louder. Increases in tinnitus loudness were associated
with increased activity in right auditory cortex whereas decreases were
associated with decreased activity in auditory cortex. Collectively, these three studies suggest that tinnitus arises from aberrant in the central auditory brain system brought about by the loss of input from the cochlea.
To investigate the mechanisms of tinnitus in more detail, we developed a behavioral technique, schedule induce polydipsia avoidance conditioning (SIPAC) to measure to salicylate, quinine and noise induced tinnitus in individual rats. Reliable measurements of tinnitus-like behavior can be obtained from individual subjects over many weeks or months. All rats reliably developed behavioral evidence of tinnitus when treated with high doses of salicylate and quinine, but not with low doses or placebo. The tinnitus-like behavior disappeared after 1-2 days after terminating these treatments.

Memantine (Namenda), an antiglutamatergic drug that acts on NMDA receptors, and scopolamine (SC), an anti-cholinergic drug, have been proposed as possible pharmacologic treatments for tinnitus. To determine if memantine and scopolamine could block the effects of salicylate-induced tinnitus, we tested the ability of memantine and scopolamine to suppress salicylate-induced tinnitus using the SIPAC paradigm. Neither memantine nor scopolamine completely suppressed salicylate induced tinnitus at drug doses Exposure to high intensity noise is one of the most common causes of tinnitus in humans; however, not all humans exposed to intense sounds develop tinnitus. When we unilaterally exposed rats to high-level noise, some rats developed behavioral evidence of tinnitus after the noise exposure; in some cases the tinnitus-like behavior pesisted, but in other cases, it disappeared after a day or two. Some rats never developed tinnitus.

SIPAC is an extremely useful behavioral paradigm for assessing tinnitus; however, it takes approximately three weeks to train an animal. To speed up our behavioral assay to assess tinnitus, we implemented a second paradigm, gap pre-pulse inhibition of acoustic startle (GPIAS). We evaluated salicylate-induced tinnitus in rats that had been tested with both SIPAC and GPIAS paradigms. Rats that developed salicylate-induced behavior with the SIPAC paradigm also showed tinnitus-like behavior with GPIAS. The concordance between the methods helps to establish
the validity of SIPAC and GPIAS paradigms.

To identify the neural correlates of tinnitus, 16-channel, chronic microwire electrodes arrays were implanted in the auditory cortex to measures single unit or local field potentials from awake rats before and after treatment with high doses of salicylate that reliably induce tinnitus. The local field potentials from the auditory cortex increased in amplitude after salicylate, especially at high frequencies. In preliminary studies, high doses of salicylate unexpectedly caused a significant decrease in spontaneous activity in awake rats. To obtain an overview of the activity in the rat brain during salicylate-induced tinnitus, we used a microPET camera and fluoro-deoxyglucose tracer (FDG) to study the changes in metabolic activity in rats with salicylate-induced tinnitus. Uptake of
FDG tracer into the brain was measured in the same rat under normal baseline conditions and after the induction of tinnitus with 250 mg/kg of salicylate. During salicylate induced tinnitus, inferior colliculi (P=0.03) and auditory cortices (P=0.003) showed significant increase in FDG activities, whereas there was no significant difference in thalamic activity (P=0.07) from the pre-salicylate, baseline state. Subjective tinnitus was once considered a difficult if not impossible problem to investigate. However, over the past two decade, rapid advances in brain imaging, behavioral models and techniques for recording from awake subjects experiencing
tinnitus have advanced our understanding of the biological mechanisms that may be involved in tinnitus. Animal models have the potential to reveal the biological mechanisms that give rise to tinnitus and offer the potential to screen existing or new therapeutic compounds that may suppress tinnitus.

Acknowledgments: Research supported in part by grants from the American Tinnitus Association and Tinnitus Research Consortium.