Tinnitus perception has been hypothesized to arise partly from spontaneous neural hyperactivity in the central auditory system. Benzodiazepines, the positive allosteric modulators of the GABAA receptor are expected to potentiate the inhibitory neurotransmission and lessen tinnitus symptoms by reducing this hyperactivity. As they have also beneficial effects on comorbid anxiety and insomnia which are common in tinnitus patients, there remains the possibility that the effect of these drugs is the result of a general anxiolytic effect rather than a direct effect on the neurophysiological cause of tinnitus [5]. Six clinical trials of benzodiazepines revealed that benzodiazepine use for subjective tinnitus does not have a robust evidence base [6] (Table 2). Clonazepam has the most evidence to support its use [789]. The effectiveness of alprazolam is equivocal [1011]. Diazepam had no effect and oxazepam was effective in only one study [712]. Benzodiazepines carry a risk of misuse or abuse and have a considerable list of adverse effects such as sedation, drowsiness, memory impairment, and unsteadiness. Clonazepam is relatively less likely to lead to abuse because of its longer half-life [6].

Antidepressants have been frequently used for the management of tinnitus because of a high comorbidity of tinnitus and depression. It has been long disputed whether tinnitus is more likely to occur in psychologically disturbed people or whether tinnitus causes the psychological disturbance. There is also debate about whether psychoactive drugs act on the central auditory system to reduce tinnitus directly or whether they act by treating comorbid depression. It has been hypothesized that some drugs may have direct effect on tinnitus since many receptors on which psychoactive drugs operate are also present within the central auditory pathways [13]. Also, it should be considered in the assessment of antidepressant effect on tinnitus in that the scales for tinnitus severity correlate highly with the depression scales [6]. In all clinical trials of tricyclic antidepressants (trimipramine, nortriptyline, and amitriptyline), there was no evidence that those drugs are effective or ineffective in the management of tinnitus although there was a slight improvement in tinnitus due to methodological bias [14151617]. The trials of selective serotonin reuptake inhibitors (paroxetine, sertraline, and trazodone) showed no benefit for most outcome measures in non-depressed patients [181920] (Table 3). Nevertheless, the use of an antidepressant in tinnitus patients with depression and anxiety is highly indicated because this patient group can be benefited from antidepressant treatment. Overall, there is no high-quality evidence on the effect of antidepressants on tinnitus reduction. The common side effects include sedation and dry mouth.

Anticonvulsants, i.e. antiepileptics are increasingly used in sev-eral non-epileptic conditions including psychiatric disorders and pain syndromes. They also represent potential candidates for the treatment of tinnitus on the assumption that tinnitus is related to central auditory hyperactivity [6]. Anticonvulsants reduce neuronal hyperexcitability through three main pharmacological mech-anisms of action: halting depolarization by blocking voltage-dependent sodium channels, augmenting GABA action, and lessening glutamate transmission [21]. Carbamazepine reduces neural firing by binding to voltage-gated sodium channels and stabilizes the neuronal membrane. A significant benefit from carbamazepine has not been confirmed in placebo-controlled studies [2223]. Gabapentin is an anticonvulsant that mimics the chemical structure of GABA, but does not acts on GABA receptors. Although the exact mechanisms of action have not been fully described, gabapentin is thought to increase GABA biosynthesis and to interact with voltage-gated calcium channels to reduce calcium current after chronic application [2425]. The side effects of gabapentin are relatively less than for other GABAergic drugs. There is insufficient evidence supporting the effectiveness of gabapentin for tinnitus [26272829]. Lamotrigine that suppresses glutamate release and inhibits voltage-dependent sodium channel failed to demonstrate a beneficial effect on tinnitus [30]. Valproic acid, one of the most frequently prescribed antiepileptics has not been systematically investigated for tinnitus. Anticonvulsant studies performed so far only showed small effects of doubtful clinical significance and there is no evidence for a large positive effect of anticonvulsants in the treatment of tinnitus [621] (Table 4). Side effects are experienced by a relatively large portion of the patients (18%) [21]. Retigabine, an approved novel antiepileptic drug, is a KCNQ2-5 channel activator. Voltage-gated Kv7 (KCNQ) channels are potassium channels activated at resting membrane potentials, and provide a powerful brake on neuronal excitability. Recent research on KCNQ2/3 channel activators have produced promising results in preventing the development of tinnitus in mouse models [3132]. It has been reported that a reduction in KCNQ2/3 channel activity results in tinnitus-specific hyperactivity in the dorsal cochlear nucleus and initiates the development of tinnitus. A pharmacological manipulation that shifts the voltage dependence of Kv7 to more negative voltages prevented the development of tinnitus after noise exposure [32]. Another KCNQ2/3 channel activator, a more potent and less toxic than retigabine in rodents, has been shown to reduce the development of tinnitus in the noise-exposed mouse model of tinnitus, providing a clinical candidate for preventing tinnitus [31].

Glutamate is a major excitatory neurotransmitter within the cochlea and central auditory pathways. Disturbances of glutamatergic transmission have been implicated in various disorders of the central nervous system and also in the development of tinnitus. These alterations may change the balance between the excitatory and inhibitory neurotransmission in the central auditory pathways. The NMDA receptor (NMDAR) is a specific type of ionotropic glutamate receptor, named after the agonist molecule N-methyl-D-aspartate that binds selectively to it. The NMDARs projecting from the medial geniculate body to the amygdala are hypothesized to be of importance in a pathophysiological model of tinnitus, in which the connections between the auditory, limbic, and autonomic nervous systems play a crucial part in the emergence of tinnitus [33]. Caroverine, an antagonist of non-NMDA and NMDA receptors, has shown contradictory results [3435]. Acamprosate, the putative non-selective NMDAR antagonist and GABA agonist, which is approved for the treatment of alcohol dependency, resulted in significant improvement of tinnitus [3637]. The non-selective NMDA antagonist memantine was no more effective than placebo [38]. Recent research demonstrated that memantine significantly reduced tinnitus-like behaviors caused by salicylate or acoustic trauma in rats [3940]. The memantine analog neramexane, which blocks both NMDA and α9α10 nicotinic cholinergic receptors, showed a trend towards improvement of tinnitus in the medium- and high-dose neramexane groups in a Phase II study. However, several Phase III studies have reported no sufficient evidence for approval of the drug [6] (Table 5).

The dopaminergic pathway which includes the prefrontal area, primary temporal area, temporo-parietal associative area, and limbic system has been proposed as the structure that supports the neurophysiological model of tinnitus perception. The cerebral areas where tinnitus is perceived are the identical areas where dopa-mine is localized in. The studies using imaging techniques have identified three zones where tinnitus is perceived: the prefrontal area associated with attention and stress; the primary temporal area in relation to hearing; and the limbic system that controls emotions, learning, memory, and motivated behavior [41]. Increased activity in the hypothesized auditolimbic dopaminergic pathway is potentially involved in emotional aspects of tinnitus and may produce the distress in tinnitus sufferers. This theory provides a means to modulate the dopamine actions using dopaminergic agonistic and antagonistic drugs [42]. Sulpiride, a dopamine D2 antagonist and atypical antipsychotic drug, significantly reduced tinnitus perception, modulating the auditolimbic dopaminergic pathway [42]. Piribedil, a dopamine D2/D3 agonist approved for Parkinsonism, was not superior to placebo. Instead, the incidence of side effects such as nausea and dizziness was relatively high, resulting in treatment interruption in all cases [43]. Another D2/D3 agonist, pramipexole, produced a beneficial effect on psychoacoustic measures of tinnitus in presbycusis patients in a dose schedule accepted for the treatment of Parkinson's disease in elderly people. A decrease in D2/D3 receptor density and function in advanced age may contribute to the onset and persistence of tinnitus [44]. The dopamine concept of tinnitus perception has opened possibilities for dopamine supplementation or receptor modulation using agonists or antagonists, weighing the agonists more strongly.

Baclofen, a GABAB agonist with muscle relaxant effects, was not more effective than placebo in a clinical trial in which a clinically available racemic baclofen was probably used [45]. However, in rats with acoustic trauma, a reduction of tinnitus behavior has been observed after application of L-baclofen [46]. L-baclofen has been reported to be more potent than both D-baclofen and racemic baclofen in reducing sound-evoked hyperexcitability in neurons of the inferior colliculus, and in relieving the symptoms of trigeminal neuralgia with better tolerated side effects [47]. Cyclobenzaprine, a muscle relaxant that modulates muscle tension, has been used to relieve skeletal muscle spasms and associated pain in acute musculoskeletal conditions. It is known to be a tricyclic antidepressant analog which inhibits the uptake of norepinephrine, resulting in increased synaptic norepinephrine concentration [48]. In a pilot study without a placebo, high-dose cyclobenzaprine resulted in a significant reduction in the tinnitus handicap inventory score [49].

Ginkgo biloba extract has been widely prescribed in peripheral vascular disease and cerebral insufficiency. The standardized extract of ginkgo biloba should contain the active chemical components of 24-25% flavonoids and 6% terpenoid fraction containing ginkgolides and bilobalide. The flavonoids have antioxidant properties, while ginkgolide B has potent PAF antagonism. Many of the central nervous system effects of the extract have been dependent on the combination of its antioxidant and PAF antagonistic actions. It is also a vasodilator, which may be the obvious reason for thinking that it would be useful in the management of tinnitus [50]. However, two most systematic double-blind and placebo controlled clinical trials have yielded negative results suggesting that ginkgo biloba extracts are of little more use in the treatment of tinnitus than a placebo [5152]. Melatonin synthesized by the pineal gland is an endogenous antioxidant, which also seems to affect sleep, mood, reproduction, immunity, and aging. Melatonin treatment showed beneficial effect on tinnitus, especially in the subgroup of patients complaining of sleep disturbance. Melatonin alone or in combination with other drugs could offer useful perspectives in the management of patients suffering from tinnitus and sleep disturbance [53].