While there is no evidence for nearly all direct procedures that intend modulating or stimulating either the cochlea or specific cervical regions such as the auditory cortex, there are therapeutic procedures that are acknowledged in clinical practice and have achieved at least a certain degree of evidence and generate measurable effect sizes. Those are in particular habituation therapy and psychotherapeutic measures, especially if they are combined with concrete measures for improved audio perception (hearing aids, CI, hearing therapies).

A nearly endless number of procedures has been tried and in particular sold for the treatment of tinnitus, unfortunately they have not been evaluated appropriately in an evidence-based way. A causal therapy, omitting the tinnitus still does not exist, actually it cannot exist because of the various mechanisms of its origin. However or perhaps because of that, medical interventions appear and reappear like fashion trends that can never be proven by stable and reliable treatment success. This contribution will discuss and acknowledge all current therapeutic procedures and the existing or non-existing evidence will be assessed. Beside external evidence, the term of evidence also encompasses the internal evidence, i.e. the experience of the treating physician and the patient’s needs shall be included.

If this situation that lasts already for a long time really leads to evidence gaps or if this gap can be partly closed by clinical expertise, remains to be observed and finally also to be discussed in the last part of this chapter.

Sometimes, in particular in the context of publications from China and Korea, only the abstract is available in English, the trial itself is published for example in Chinese language. Of course, this explains why so many studies are not considered for meta-analyses.

Nonetheless, an extremely high number of studies is published even if only very little data is present which furthermore is not even reliable. In many countries, research and even part of the clinical work is financed via funding, hereby publications are essential to have access to funding. Additionally, numerous online platforms facilitate publication. Potential authors are “invited” to submit studies, rapid processing is promised. The authors have to co-finance these kinds of publications: often, however, a peer-reviewing does not work soundly. Furthermore, the high number of submitted articles worldwide causes a certain tiredness among the reviewers. If someone has to check the data of a trial thoroughly, this needs time and learning the matter. So it can be expected that also the reviewers only marginally verify many studies. This aspect is obvious regarding the fact that even some high-quality journals (with high impact factors) often accept trials that are methodically false or show significant flaws and do not even meet the basic criteria of scientific working and ethical obligations. This also means that the basic rule of anonymous submission is often neglected – certain authors may then publish nearly everything, regardless of actual quality and scientific and especially clinical relevance. Due to the internet, today nearly all studies are available on a worldwide scale, most of them at least via the abstracts. However, those abstracts are often too short and especially in the context of therapy studies they sometimes describe conclusions that actually cannot be drawn from the trial and are not even mentioned explicitly. Since many authors only read the abstracts for their discussion chapters, systematic errors are included and distributed.

Finally it is important to mention that the conduction of large, multicenter, and high-quality studies causes significant costs. Since in general pharmaceutical companies are not available as sponsors (because success cannot be expected) for tinnitus therapy studies and state funding is rather reluctant, those are other reasons explaining why there are only few reliable trials in this context.

The authors refer in particular to pain therapy that can certainly be compared to the therapy of chronic tinnitus and criticize that often clinically proven therapies are doubted or not approved. This becomes obvious regarding the fact that according to El Dib et al. [29] 96% of a total of 1024 review articles do not give a definite statement on existing or missing evidence.

For studies on tinnitus therapy, especially those homogenous research objects are rarely found because in particular chronic tinnitus is influenced and characterized by numerous factors that are relevant for the actual appearance and suffering from the symptom. So, the research objects are rather heterogenic which is a main criticism according to Feinstein (cited after Weßling [25] ) in the sense that such heterogenic characteristics cannot be summarized satisfactorily in a meta-analysis. Furthermore, trials with negative therapeutic effects are generally less frequently published (publication bias). In 2008, Turner et al. [26] could well show this aspect based on an investigation of the efficacy of antidepressant medication. They elaborated that numerous registered studies have not been published while it is not clear if they had not been accepted for publication or not submitted at all. Subsequent meta-analyses thus often have false positive assessments. This fact mainly concern studies and meta-analyses that are sponsored by pharmaceutical companies [27] . But if – such as often in the context of trials on tinnitus therapy – only little proof and comparable studies exist, clear evidence and good reasoning is nearly not possible: criteria of evidence-based medicine are not fulfilled although some therapeutic approaches might be effective. According practice guidelines that only refer to the basis of evidence-based medicine thus cannot recommend those therapeutic approaches. This fact is well elaborated in a publication by Kern et al. [28] (describing the example of physical medicine and general rehabilitation): The authors explain that the original definition of evidence-based medicine included individual clinical experiences beside external evidence, however, that this gets increasingly in the background and only external evidence is considered as the most important guideline for financial compensation of therapies and therapeutic guidelines.

Nowadays, the bases for evidence-based medicine are so-called meta-analyses representing a summary of primary investigations that are evaluated in a qualitative and statistical way. For such meta-analyses, different trials in a research field are studied, then summarized, and statistically evaluated. The investigations primarily included should preferably be homogenous in order to allow consistent data collection and then to analyze them in a statistically sufficient way.

According to Coyle [22] , the preconditions for external evidence are fulfilled only in well-structured and high-quality studies conducted in a randomized, controlled, double-blind, and generally also placebo-controlled way with sufficiently high numbers of patients. Those investigations are classified according to validity criteria into more or less strong evidence.

During the last years, evidence-based medicine turned out to be an independent science which is the gold standard especially for evaluation of therapies and therapeutic recommendations even if also this aspect is not without controversy. Evidence-based medicine is built on three pillars:

In summary, in the discussion of scientific investigations, the current preconditions for evidence determination are still not a really reliable tool to differentiate effective from non-effective measures. Hence, the “old” requirement remains [21] that probably only networking and central assessment of many different therapeutic approaches and centers may provide reliable data.

In particular, cognitive behavioral therapies – especially when manualized – can be very well assessed and thus also standardized. For according meta-analyses this is always a very important criterion leading to the fact that other comparably effective (also psychotherapeutic) therapy approaches do not reach the same degree of evidence.

However, the meta-analysis of Martinez-Devesa [16] for example and other studies with limited study design must not lead to the claim of superiority of “cognitive behavioral therapy” for outpatient individual therapy or in daily routine [17] .

Even if often special conditions are valid for such trials, e.g. because only less stressed patients were included, the significance is not reduced regarding the statement that the evaluated elements from cognitive behavioral therapy are effective in patients suffering from tinnitus. So they are – as published in the manuals by Kröner-Herwig [14] and Delb [15] – useful parts of each serious tinnitus therapy.

Only very few trials meet the requirement that is the base of medical evidence. Actually the question must be asked if trials included in according meta-analyses really meet the criteria without simply omitting certain therapeutic aspects, as described above. Better evidence is found for trials on the treatment of chronic tinnitus that (nearly always) concern cognitive behavioral therapy.

From experience, that scientific assessment of such therapies and devices is very difficult, sometimes trials are mentioned in single cases, but also in this context the possibilities of control and especially the placebo-controls are extremely difficult if not even impossible. For example, even for non-experts a placebo coil is recognized as not being active in the context of transcranial magnetic stimulation. The same applies for tonal stimulation that then places stimuli at totally different locations than corresponding to the own tinnitus.

Even more problematic are therapies that are based on technical solutions such as for example radiation or acoustic alienation. In this context, often technical inspection certificates are mentioned as quality proof that certify a general possibility to be used in a medical context to a device (especially because it does not cause direct harm), however, no clear statement can be given on the therapeutic success and the actual benefit for the patient. While reliable studies in three clinical phases are required for pharmaceutics, this rule is not applied to technical devices. So nearly every device that does not cause damage can be sold as therapeutic instrument and promoted as such.

Generally, studies on the efficacy of pharmaceutics are only rarely placebo-controlled, often one group of substances is compared to another which neglects the placebo effect that is extremely high in the context of tinnitus [13] . The scientific importance is thus clearly reduced. Often pharmaceutical or instrument-related therapies are performed in ENT practices as so-called observational trials – without control and validated evaluation. From the start, those studies are useless, nonetheless they are often taken as basis for commercial advertising.

In reality, many therapies and approaches evaluating single measures scientifically are always successful with combined therapeutic modules such as the training of relaxing techniques or psychological counseling and stabilization, often even psychotherapy in scientifically sound procedures. Then often only one therapeutic element is in the focus, the others are “forgotten” in order to relate therapeutic success to only one treatment. However, from a scientific point of view this is not “neat” and falsifies the data situation.

All this means that tinnitus therapies – if they are expected to have positive results – never work as mono-therapies, especially not under time-restricted conditions of regular ENT practices that only rarely allow sufficient information. However, this fact makes it problematic to scientifically evaluate solid and valid trials because the single therapeutic effects cannot be clearly related to the one or other measure. So it is evident that good and successful tinnitus treatment must be combined with compensation of parallel and accompanying hearing loss, in general by prescribing hearing aids. Up to now, the evidence that hearing aids are useful for the treatment of tinnitus could not be proven because this treatment is always only part of the therapy and cannot be assessed separately. Thus, those trials are not considered in meta-analyses.

Only very rarely, when the hearing capacity is completely regular, also after measuring OAEs, the ear noise will have to be understood as a consequence of general over-stimulation and thus incorrect processing of the central auricular system.

In terms of diagnostics, it is necessary to identify exactly possible hearing impairment even if the patient himself considers his hearing capacities as regular. Beside the pure tone audiogram, in particular the assessment of the function of the external hair cells is reasonable and required by measuring DPOAE.

Thus, tinnitus corresponds to a primary functional loss of mainly external hair cells of the inner ear as described recently by Noreña [8] in a review article. However, mostly enhancement mechanisms in the auricular system are responsible for the perception of tinnitus that enhance this frequency via cortical reactions and priming or reduction of lateral inhibition or increase of base frequencies so that the impression is dominant. This means that a peripheral dysfunction leads to central tinnitus enhancement or tinnitus accentuation. In accordance, often the distortion products of otoacoustic emissions are nearly paradoxically increased [9] , [10] .

In the majority of the cases, tinnitus is associated with hearing loss. Sometimes this hearing loss is subjectively not perceived at all or as annoying effect especially when it has slowly developed. According to own data, tinnitus patients only rarely have regular hearing capacities [7] . This means that tinnitus is mostly observed in the frequency of the highest hearing loss, mostly imposing as high-frequency sound because of the dominance of high-frequency hearing impairment. Sudden hearing impairment as for example in the context of sudden hearing loss or noise trauma is often accompanied by tinnitus that often occurs only in the interval, i.e. when the hearing impairment improves or remains on a certain level.

Based on the cortical network, especially the combination with emotional aspects and negative assessment, annoying ear noise develops and as a consequence also comorbidities such as concentration and sleep disorders, partly even depressive episodes and anxiety. Those comorbidities are actually significant for the development of an independent disease; they lead to the fact that the tinnitus aurium must be considered as complex disease, especially in cases of patients requiring treatment. In terms of therapy, this has a significant impact because only mechanistic, symptom-related approaches have little prospect of success due to this complexity.

Even for acutely arising hearing disorders, tinnitus is observed often only in phases of relaxing or after a longer interval. In contrast, tinnitus is suddenly subjectively perceived in the context of slowly developing deafness that occurs in the majority of the cases, often triggered by stress or emotional distress. Because of this close relationship with the hearing perception, tinnitus aurium is not considered as an independent disease but as a symptom of disturbed hearing at any location of the auricular system.

93% of all tinnitus patients have concomitant (doubtfully causing?) measurable hearing loss, 44% complain also of hyperacusis. Most ear noises are observed in the high frequencies and impose as high whistling sound; they nearly always correspond to a simultaneously existing hearing loss in those high frequencies [5] , [6] .

The pure disease-related costs for healthcare services are lower compared to the expenses that are caused by social side effects (inability to work, early retirement etc.). In terms of therapy of tinnitus aurium, it is important if its occurrence is acute or already present for a longer time so that it must be considered as chronic. Whereas in the acute stage it is recommended to initiate high-dose cortisone therapy (with moderate but however existing evidence) in analogy to the treatment of sudden hearing loss, there is no known treatment for chronic tinnitus that might causally stop the ear noise.

In rare cases, tinnitus can be objectified, it is then described as a pulsating noise or a clicking or smacking and corresponds rather to perceived sound produced by the body such as vascular processes or muscle contractions. Much more frequently, subjective tinnitus is observed that cannot be assessed by external measurement and that is caused in more than 90% by a dysfunction of the hair cells in the inner ear and processed as annoying sound [2] . According to recent studies [3] , about 15% of all Chinese suffer from tinnitus while the prevalence in higher ages is increased. The main risk factors are hearing loss, middle ear affection, and noise exposure. This leads to significant costs due to the high percentage of patients requiring treatment, mostly for the patients themselves, but also for the healthcare insurances. According to an evaluation of the disease-related expenses [4] , the costs amount to nearly 7 billion Euro only in the Netherlands.

Tinnitus (from the Latin word “tinnire” = ring) is synonymous for ear noise or ear ringing and corresponds to the perception of an acoustic phenomenon that is not caused by external impulses. It is generated at some point of the auditory system, mostly in the cochlea, then it is processed and perceived as noise annoyance in the cortex. Such as all recurrent continuous stimulations, it is habituated in most of the cases by sensory perception, i.e. filtered already subcortically. If this is not possible or if particular attention is paid to the tinnitus, it may sometimes lead to annoyance and also subsequent complaints causing separate disease. Epidemiologic studies for Europe and the United States of America, expect about ¼ of all people having experienced tinnitus sensation at least once, while 10–15% hear tinnitus for a longer interval, actually only 3–5% are considered as requiring treatment, and half of them suffer significantly [1] .

Practice guidelines may be very helpful for treatment especially when – as in the USA – therapies are classified directly as not being useful. In Germany, there is no such negative valuation (partly because of fear of prosecution) and only therapies are recommended in the guideline for which a certain evidence is confirmed. Of course this also means that other approaches, in particular the many and partly very expensive instrument-based procedures, cannot be recommended.

Also in this context, pharmacotherapy is not recommended, neither alternative therapies or oxygen therapies or magnetic radiations; the recommendation to hearing aids, cochlear implantation and hearing or sound therapies remains open according to this guideline because there is no sufficient evidence. Recommended as evidence-based is the treatment of comorbidities, especially anxiety and depression.

In contrast, the German S3 Guideline: Tinnitus from 2015 [31] reports about confirmed evidence of tinnitus therapies. According to this document, only tinnitus-related counseling and cognitive behavioral therapy can be recommended as evidence-based treatment. However, this guideline generally recommends cognitive behavioral therapy and not only “manualized tinnitus-related cognitive behavioral therapy”, as Zenner [41] erroneously writes in his summary for the journal HNO.

In this US-American guideline, some therapies are explicitly marked as “recommendation against” because there are no reliable study data. Therapies for which the evidence of existing trials is not sufficient, are described at least as “optional” as for example hearing and sound therapy.

Since 2014, there is a “Clinical Practice Guideline: Tinnitus” [39] ; it contains definitions and a total of 13 therapy recommendations. An “executive summary” [40] explains the methods of the recommendations and suggests evaluation studies and further evaluations.

In this article, single therapeutic approaches for acute and chronic tinnitus will be presented and valued, with regard to possible or also missing evidence and also with regard to a relevant therapeutic significance for the ENT specialist.

Hence, even numerous trials often cannot confirm an evidence of the according therapy. But if classic evidence in the sense of external evidence is not necessarily required for effective tinnitus therapy, many of the presented studies of more clinically oriented assessment of the efficacy lead to no or even negative significance.

Those preliminary reflections shall only give an impression how tense the field of tinnitus therapies is and which expectations patients as well as therapists often have regarding short-term relief. However, evaluating more intensively the multiple possibilities of the origin of ear noises, especially the development of real and clinically assessable distress for the patient, it is clear that mechanistic proposals that focus only on one aspect have to be directly excluded. But this does not happen because such mechanistic approaches are pursued due to different interests. Frequently the results are methodologically “cleaned up”. Sometimes even secondary or tertiary effects are first assessed in so-called post-hoc analyses that identify an effect even if primarily no effect can be confirmed as for example in subgroups of hypertension patients or patients suffering from tinnitus caused by middle ear disease.

In a subsequent trial, 439 patients were asked again: 40% had already spent between 500 $ and 10,000 $. 70% would accept implantation of a device or stimulator to just reduce the distress caused by tinnitus [11] .

In parallel, there are especially the patients who – if bothered by tinnitus – require preferably causal therapy stopping the tinnitus. Also in this regard there are interesting studies that focus on the actual readiness of the patients to undergo certain therapeutic options. Rich Tyler from Iowa dealt with the questions what patients would be ready to tolerate to make the tinnitus more bearable [12] . According to this trial, at least 19% of the examined 197 patients would accept an implant in the brain if they completely lost their tinnitus. 13% would even accept such a treatment if the tinnitus was half as loud. The readiness to take pills amounted to more than 50% for both questions. Most patient would spend up to 5,000 $ to lose their tinnitus, 20.3% would even spend more than 25,000 $.

Also in 2011, a review article on tinnitus therapy was presented in a neurologic journal that could not have been more superficial [37] : For therapy of chronic tinnitus numerous reasonable but also inappropriate, non-confirmed therapies without scientific assessment were presented. According to this review article, nearly everything seems to be effective, starting with hearing aids, noisers, and cochlear implants, but also pharmaceutics such as gingko and caroverine, low-power laser therapy up to cognitive behavioral therapy. If therapies are compared in such a confuse way and therapeutic success is reported even in review articles without any comment, it does not help finding useful therapeutic decisions. In many review articles and meta-analyses the criticism is found that there is none of the common treatment approaches for chronic tinnitus working as monotherapy, generally multimodal approaches are applied and then evaluated. However, the scientific assessment for the single therapeutic options becomes difficult and even impossible. This is also true for the often emphasized cognitive behavioral therapy because this treatment is nearly always combined with other procedures as for example hearing therapies or relaxation exercises. In general, those are not included in the evaluation and quasi concealed. Additionally, blinding or even placebo therapy for those therapeutic approaches is impossible. In 2012, a multidisciplinary study group tried to make a methodological proposal for improved studies [38]. Even this proposal of an international standard for tinnitus therapy correctly states some of the problems, but it does not really consider the main problems of scientific evaluation of tinnitus therapies:

In general, the methods of studies on the efficacy of cognitive and neuro-otologic psychosomatic therapies improve continuously while trials on the direct influence of the tinnitus – either by pharmaceutics or by cortical modulations – are often extremely superficial and methodologically rather poor. Then too readily successes are announced that have to be withdrawn shortly afterwards or the according therapies have already disappeared from the market. However, this situation has been observed for tinnitus therapy for more than 40 years. An interesting review article from the USA was published in 2013 [35] . Via internet, more than 9000 trials and publications between 1970 and 2012 were assessed, but only 52 could be evaluated in terms of therapeutic effects in cases of tinnitus. Among those publications, 17 evaluated pharmacological therapies, 11 dealt with other interventions such as TMS or laser treatment, 5 with sound therapy, and 19 with approaches of psychological behavioral therapy. The authors complain about the aspect that nearly no data is given on side effects. In total, they found only weak evidence for the efficacy of cognitive behavioral therapy with regards to tinnitus-related improvement of the quality of life. Weak evidence was further seen regarding the loudness of the tinnitus for the efficacy of neurotransmitters compared to placebo. Insufficient evidence was found for the efficacy of antidepressants, other pharmaceutics, and food supplements regarding loudness of the tinnitus and all other therapeutic goals. Insufficient evidence was also identified for acoustic neuro-stimulation, rTMS, and sound therapies, partly the studies had a high bias, for example caused by direct economic involvement of the authors in selling the evaluated medical devices. But this review article also mixed up possible therapeutic effects and did not differentiate if the topic was the distress caused by the tinnitus (which can be assessed by specific questionnaires) or direct audiological factors such as loudness and frequency of the tinnitus. A meta-analysis from Nottingham submitted in 2011 [36] considered 28 randomized and controlled studies on tinnitus therapy and addressed crucial subjects: many of those trials had poor evidence because they had not been blinded, their significance was not well measured, and additionally the data was often not completely reported. Also in this context, those were only trials on cognitive behavioral therapy that were sufficiently large and thus comparable and that could assign a certain – even if modest – effect size to this type of therapy. In this evaluation, a certain evidence for antidepressants was found, however, it was not clear of what the therapeutic effect of antidepressant for tinnitus really consisted.

For therapeutic reflections, it is crucial to evaluate the actual distress caused by the ear noise and the developing side effects and comorbidities. If an ear noise does not really disturb the patient and if it is not made responsible for other disorders, therapy is not really required. The ear noise will then be suppressed simply by habituation processes. The basic knowledge about the significance of the chronic tinnitus and its assessment as a disease still remains that the actual persistence of tinnitus and consequently the developing distress are only due to cortical plasticity [8] , [34] . Even if originally a hearing damage, especially damage of the external hair cells of the inner ear is present, the processing in the auricular system and the interconnection with emotional and evaluation qualities in the auricular system that are basic for the significance and the pathology of this ear noise. In a certain way, also the conflict is resolved hereby that came up during the last years with regard to the genesis of ear noises. Especially we as ENT specialists expect a primarily peripheral genesis in the hair cells of the inner ear while e.g. neurologists and psychiatrists indicate preferably the central significance and erroneously state that tinnitus is centrally generated. It is correct in this context that distress caused by tinnitus develops and is generated centrally, however, the ear noise itself develops at different locations and mostly really in the inner ear. This is important for therapeutic options that have to take into consideration the primary genesis as well as the further central processing.

However, it is important to differentiate between acute and chronic tinnitus because according to the current study situation and to the actualized guidelines pharmacotherapy should only be attempted in the really acute stage whereas it is no longer useful in cases of chronic tinnitus.

In cases of objective or objectified ear noises that are actually very rare, a basic pathophysiology exists that may even be treated surgically. For example pulsating ear noise may be caused by arterio-venous fistulas or vascular processes. In this context, it must be discussed if surgery is appropriate and necessary, which depends on the degree of real tinnitus severity and also of probably resulting comorbidities.

Another frequently occurring comorbidity is anxiety caused by the fact that tinnitus is seen as threatening symptom and fear of aggravation or new damage develops. Often it is associated with a hypersensitivity to noises (hyperacusis).

More or less frequently, this increasing focusing leads to psychosomatic comorbidities such as sleep and concentration disorders. Primarily tinnitus is made responsible for them because the patient can no longer concentrate due to the tinnitus, can no longer sleep due to the tinnitus, and even wakes up due to the tinnitus. Often social isolation and depressive reactions develop and again tinnitus is made responsible for this situation, although certainly also accompanying exhaustion, stress, and a general depression may be the origin.

Finally, it is decisive for therapy and the need of therapy of the tinnitus patient how he is able to cope with the phenomenon and to what extent he suffers. If we expect that more than 25% of all Germans have already experienced tinnitus but only 2% feel impaired by the ear noises, there are numerous patients who rapidly accept their tinnitus as given and habituate. The ear noise does not cause an urgent need of therapy and is not perceived as annoying. Therefore they do not hear it permanently. Regular habituation processes in the hearing processing compensate the tinnitus, completely independent from the timely phase or the duration of its presence. However, if the impression of tinnitus is combined and enhanced in the brain by plastic alterations and linking in the emotional assessment and consecutively by focusing reactions, the regular habituation is impeded and suffering from the tinnitus results [32] . Tinnitus tends to decompensate or already decompensates the patient. The symptom then dominates the affected person, controls his ability to live and decide and impairs it more or less intensively [5] , [33] . This possible development is completely independent from the circumstance if the tinnitus was generated primarily in the inner ear, in the hearing nerve, the brain stem, or the central hearing processing.

Hereby it is important that even in the context of chronic tinnitus of longer durations there are situations when tinnitus becomes louder or more intensive. This aggravation which is commonly called exacerbation, however, cannot be compared to acutely occurring tinnitus. It is rather only a perception phenomenon that is triggered by certain situations such as stress.

Especially with regard to therapeutic interventions, the duration and persistence of the ear noise are highly relevant. An acute tinnitus occurring for the first time often disappears spontaneously after a short time or after according therapy. Only if the ear noise persists for more than 3 months, it is called chronic. This also depends enormously from the attention the patient pays to the phenomenon. The classification of acute and chronic is important because it defines if acute therapies is still successful in this phase of the disease. According to the common guidelines and therapeutic experiences, an acute therapy is only useful within the first 3 months [30] , [31] .

Beside the already mentioned differentiation between objective and subjective tinnitus and tinnitus with or without hearing loss, it is important for therapy if tinnitus occurred acutely or if it is observed already for a longer time and especially if it is compensated or decompensated, which means if and to what extent already comorbidities have developed.

Patients claim from themselves and even more from their treating physicians to give understandable causal explanations about the symptom of tinnitus. Ear noises may occur in various forms, as pure sounds of different frequencies, as sound mixtures or as narrow band or broadband noise. The quality of the ear noise is highly important for the patient, however, it is not really relevant for the pathological correlations. Generally it must be said that most of the ear noises impose as highly frequent whistling. This fact also allows conclusions on the accompanying or triggering high-frequency hearing loss. In contrast, low-frequency buzzing sounds are often associated with low-frequency hearing loss and may be a hint to an endolymphatic obstruction in the inner ear. Furthermore, the tinnitus may be intermittent or permanent, its intensity may vary, at least regarding the individual perception, and may even by pulsating in single cases. For the systematic history taking it is also important to know if the ear noise can be enhanced by movements of the head, neck, or jaw or if physical movement reduces or increases the tinnitus.

In summary, there is still no sufficient evidence and no clinical experience for the effective application of pharmaceutical products in order to really and effectively treat or only suppress acute or chronic tinnitus.

In this situation, experienced physicians who can refer to complex diagnostics and a good cooperation with audiologists and psychologist are responsible for the therapeutic effect. This was described by a review article from the USA [76] .

The German as well as the American practice guidelines come to the conclusion that there is no proven effect for any medication in the context of chronic tinnitus persisting for more than 3 months. A review article from Spain confirms this statement. The literature with regard of numerous pharmacotherapeutic approaches was assessed [75] : anticonvulsants, anesthetics, antidepressants, antihistamines, benzodiazepines, corticosteroids, as well as diuretics. In single cases success is reported everynow and again but the evidence is very low. At the same time there are other trials that do not confirm any effect for the same pharmaceutical product.

Still pharmacotherapy in tinnitus disease is not very promising and has no proven evidence. Even the application of intratympanic cortisone cannot be justified for tinnitus alone and not at all for persisting tinnitus even if a certain – however rather temporary – placebo effect is observed. Zinc, ozone, and betahistine are evidently ineffective as well as ginkgo. This last aspect is very interesting and confirmed by a Cochrane analysis because especially in Germany the effect of ginkgo on tinnitus is propagated loudly, even for hearing loss and for prophylaxis of noise-induced damage it is promoted, preferably by life-long intake!

A prospective trial of 68 tinnitus patients from Turkey was performed in a randomized and prospectively controlled way [74] . In 10 sessions 27 patients received ozone as autohemotherapy twice a week, 26 patients received 48 mg betahistine per day for 3 months, and 15 patients in the control group received no therapy. Ozone is applied as anti-inflammatory agent and as complementary treatment against ischemia, betahistine is expected to have a vaso-active effect and thus improve the circulation of the inner ear. No significant differences could be revealed for none of the groups neither with regard to tinnitus severity nor to loudness.

116 tinnitus patients, all of them older than 60 years, were treated either with 50 mg of zinc per day or with placebo in a randomized, double-blind, and placebo-controlled trial. After 1 month of interruption, the groups alternated [73] . Zinc is expected to promote the postsynaptic activity of some neurotransmitters, in higher ages, the concentration of zinc in the serum is generally lower. However, this study from Iowa did not reveal a positive effect on the tinnitus after treatment with zinc in comparison to placebo.

In an older analysis, gabapentin was applied for tinnitus treatment [72] . Gabapentin is an antagonist of gamma amino butyric acid, which is an inhibitor transmitter. 52 tinnitus patients received 1,800 mg gabapentin every day for 5 weeks, the control group consisted of 24 patients who received a placebo. Both groups had a light high frequency hearing loss. After treatment, significant differences regarding tinnitus severity were not observed between both groups.

Already in 2001, an article was published in the British Medical Journal (BMJ) on the effectiveness of ginkgo extract as treatment for tinnitus in a very large patient population (1,121 participants) assessed in a double-blind and placebo-controlled trial. Ginkgo did not lead to improvement of the tinnitus penetrance and intensity, neither did a placebo [69] . Numerous trials have confirmed the ineffectiveness of the ginkgo biloba extract for tinnitus treatment during the last years, the more thorough the study was, the clearer was the result [70] . Only the tiresome application observations and paid adverts in our journals as well as one “review article” always cited by the pharmaceutical industry [71] continue reporting the contrary. However, especially this review article excluded all above-mentioned trials – the reason for this aspect is not explained in the text.

This Cochrane analysis was updated in 2013 [68] . Four new trials with more than 1500 patients were evaluated: in 3 trials, tinnitus was the main diagnosis; in the context of the 4 th study, patients were treated with ginkgo who suffered from mild to moderate dementia, some of these patients also had tinnitus. No efficacy could be proven in the patients with tinnitus alone, the dement patients with only little tinnitus distress observed a low improvement. The authors postulate that ginkgo is ineffective in patients with the main complaint of tinnitus.

Ginkgo is not effective for tinnitus treatment. In 2012, a review article from Norway as update of a Cochrane review analyzed recent randomized and placbo-controlled studies on the effectiveness of ginkgo biloba with a total of more than 6000 patients. Evidence could not be revealed neither for the efficacy in the context of cognitive deficits, dementia, apoplexy, claudication nor for tinnitus. Moreover, even if mild, several side effects were observed such as vertigo, stomach complaints, or allergic reactions, and sometimes even increased bleeding tendency [67] .

In cases of according comorbidity, antidepressants are a useful completion of the therapy and lead neither to addiction nor do they sedate the patient in contrast to benzodiazepines, barbiturates, or anticonvulsants.

In an update of a Cochrane review from 2006 and again 2009, the study situation of the efficacy of antidepressants in tinnitus therapy was analyzed. 6 studies encompassing a total of 610 patients were evaluated. Only one high-quality study was found on the effect of a serotonin re-uptake inhibitor (SSRI), but it could not confirm an effect on the intensity of the tinnitus and its severity. Studies with tricyclic antidepressant were of lower quality and could not delimit the effects on anxiety and depression from tinnitus severity. In the meta-analysis the conclusion is drawn that there is still not proof that antidepressants improve tinnitus [66] . Nonetheless, antidepressants are often successfully applied in the treatment of chronic, decompensated tinnitus, however, not for improvement of the tinnitus but for treatment of accompanying depression and anxiety. If those conditions are improved, often also the tinnitus is perceived as less bothersome. So the question of the Cochrane analysis seems to be useless because nobody will receive and take an antidepressant without having symptoms of depression or at least sleep disorders.

With the mentioned high dose of this agent, numerous symptoms disappear because the patient is stunned. Not only the tinnitus gets calmer, as promised in the title of the article, but certainly the whole patient. This article was actually rejected by the European Archives of Oto-Rhino-Laryngology, however, it appeared in another journal – good scientific customs!

Because of the high risk of addiction, the administration of clonazepam is controlled by the narcotics law already after short time of intake.

In Korea, all tinnitus patients seem to be treated with benzodiazepines. In one study the spasmolytic clonazepam was randomized crossover with Ginkgo biloba (!). A total of 38 patients received this therapy (0.5 mg of clonazepam and 40 mg of Ginkgo each of them 4 times daily). Clonazepam in high doses reduced the tinnitus loudness (in 74% of the patients), the severity (79%), and also the annoyance measured in the THI (61%). Ginkgo biloba, however, did not show any effect. The conclusion of the study was that clonazepam was effective in tinnitus therapy [65] .

This study on clonazepam combined with an antidepressant is methodically misleading because a substance combination was tested so that the term of “placebo-controlled” is not applicable because it only refers to one substance. Further it is not mentioned that the benzodiazepine clonazepam has a very high potential of addiction, disturbs the sleep rhythm, and influences the learning ability so that it should not be prescribed for therapy. A positive effect as postulated by the authors in their summary, can actually not be concluded.

Clonazepam and deanxit only cause low improvement. The study on the efficacy of benzodiazepine with an antidepressant conducted by the tinnitus group from Antwerp treated a total of 35 patients in 2 groups [64] . The study design was reported as being double-blind, randomized, and placebo-controlled, but this was only partly true. All patients were treated with the benzodiazepine derivative clonazepam, they were randomly assigned to 2 groups. Additionally, all patients of the first group received the antidepressant deanxit (melitracen, a tricyclic antidepressant drug) for 3 weeks, afterward the dose was continuously reduced. Then the patients received a placebo for 3 weeks. In the second group, the sequence was reverted, again with additional clonazepam. This application was double-blind. 7 patients interrupted the study so that 28 cases could be evaluated. Only 3 patients observed an improvement of their tinnitus after melitracen, none of them after placebo. The authors exclude a positive effect of clonazepam alone because the value of depression (in the BDI) was not improved. Finally, the authors postulate that it is reasonable to combine substances that have an effect on several neuro-transmitters. In summary, a reduction of the tinnitus complaints was reported.

In a multicenter trial (Regensburg, Brasilia), this agent was analyzed more exactly and thoroughly applied in several doses and compared to other muscle relaxants. Only a high-dose application of cyclobenzaprine (30 mg) tested in 14 patients led to a reduction of the tinnitus severity in the THI. For all other agents and low-dose cyclobenzaprine, no improvement was observed. The side effects were xerostomia, sleepiness, and constipation. The authors cannot explain the potential effect of the relaxant, however, they assume that it has a comparable effect with tricyclic antidepressants due to the similarity. They emphasize that only controlled studies with higher numbers of patients may provide reliable information on the potential effect in the context of tinnitus [63] . Trials with such substance groups are really only useful if they are randomized and placebo-controlled. Otherwise it seems that only the number of publications is important, new statements are not given by those studies. Furthermore, the rate of side effects is very high. In particular, there is not pathophysiological explanation that would justify the application of this drug in chronic tinnitus.

Cyclobenzaprine is also applied as muscle relaxant for the treatment of skeletal spasms. Because of its analgesic effect, it was evaluated in 2 studies on tinnitus treatment. 65 patients were compared to 30 patients on a waiting list. 24% of the tinnitus patients had positive reactions on the relaxant that improved by 53% with regard to tinnitus intensity and 55% with regard to tinnitus induced distress [62] .

Baclofen, which is a drug from the group of muscle relaxants for treatment of spasticity in the context of spinal injuries, was applied in rats after acoustic trauma and in the animal model it led to a reduction of the tinnitus. Current trials in humans are currently not present [61] .

Intratympanic cortisone therapy for refractory tinnitus. A prospective, randomized, double-blind, and placebo-controlled trial from Korea is presented evaluating the effect of intratympanic cortisone therapy in 30 patients with bothersome tinnitus [60] . 15 patients received intratympanic cortisone injection 4 times in 2 weeks, 15 patients received saline solution. Neither with regard to tinnitus severity nor with regard to loudness, could a significant difference between both groups be revealed although both groups observed an improvement of about 30% (evaluation 4 weeks after therapy).

Furthermore, the antiemetic drug of ondansetron was examined that is expected to be effective against tinnitus and to even improve the hearing threshold, as reported by an ENT team from Teheran [59] . In a randomized, double-blind, and placebo-controlled study, 30 patients with tinnitus persisting for more than 3 months, with and without hearing loss were treated with ondansetron and compared to a placebo group of the same size. The hearing loss was reported as only moderate over all frequencies without documenting concrete results. Regarding the tinnitus, the THI (tinnitus handicap inventory), TSI (tinnitus severity index), and VAS (visual analog scale) were assessed, in addition to the Hospital Anxiety and Depression Index (HADS) and the thresholds. While VAS and THI, anxiety and depression were not significantly different in both groups, the TSI significantly improved after drug therapy. Surprisingly, the hearing and speaking thresholds improved under ondansetron application (4–16 mg/d for 4 weeks), even only of 3 dB on the average. The hearing threshold improved more in high frequency hearing impaired patients, however, it was neither documented nor specified. The authors explain the hearing improvement and the associated improvement of the tinnitus with a blockade of nicotinic acid receptors in the external hair cells and thus an improvement of the cochlear enhancement. Follow-up beyond 4 weeks did not take place. The result that antiemetic drugs may be applied for hearing improvement, is surprising. However, a summarized improvement of the hearing threshold of 3 dB over all frequencies without differentiation and documentation does not really prove an improved inner ear performance. Further it remains totally unclear why only one questionnaire shows changes and none of the others and why a positive effect of the therapy is concluded nonetheless. Again: how could such a trial be accepted by the reviewers for publication in a scientific journal?

In a multicenter trial from England [58] , a new antiemetic and anxiolytic drug (vestipitant) was tested in 24 adult tinnitus patients. The study was randomized, double-blind, placebo-controlled, and cross-over. It was conducted for 14 days, the tinnitus severity was assessed by means of the VAS and THI. In summary, no improvement under medication was observed, the intensity of the tinnitus even significantly increased.

At least a weak positive effect without side effects was documented, but the follow-up time of 4 weeks is rather short and complete disappearance of the ear noises did not occur, as in none of the other studies as well.

A review article from Italy analyzed studies evaluating the treatment of tinnitus patients with melatonin [57] . In 5 trials, therapy success could not be found with regard to tinnitus, however, the sleep disorders improved.

A review article from San Antonio [56] emphasized the positive effect of melatonin especially as protection in the context of aminoglycoside and cisplatin treatment, but also for the treatment of tinnitus.

In a prospective, randomized, double-blind, cross-over, and placebo-controlled study from Ohio, USA, 61 patients suffering from tinnitus for more than 6 months were treated with 3 mg melatonin or placebo [55] . The tinnitus severity was assessed by means of the American TSI questionnaire. The mean age was 57.8 years. After melatonin application, 57% of the patients reported improved symptoms, but also 25% of the placebo group. The better effect was achieved in male patients, in cases of bilateral tinnitus, and in patients who had not received previous treatment and who were not depressive. The same percentage of 57% observed improved sleep after melatonin application (36% after placebo). Side effects were not registered. Follow-up did not take place, the last values were assessed 4 weeks after the end of therapy. Completely disappeared tinnitus was not described.

In several centers, the tolerability and dosage of neramexane, an NMDA receptor antagonist were evaluated based on a very complex, methodologically sound, randomized, double-blind, and placebo-controlled study [54] . A total of 431 patients with chronic tinnitus persisting for 3–18 months were included in this study. The tinnitus severity was assessed with the tinnitus questionnaire TQ12, audiometric data were collected. A dosage of 50 mg as well as 75 mg could achieve a light, however not significant improvement. Significant improvement was only achieved with the dosage of 50 mg 4 weeks after the end of therapy. Thus the neramexane study showed an improvement only at a second glance, 4 weeks after the end of therapy. During treatment, the tinnitus-induced distress did not change. Often an intensive counseling and information about the findings took place in the final conversation, such a counseling itself already has a therapeutic value and might explain modest post-therapeutic improvement. The authors recommended a phase III study, also conceived as multicenter trial, but it was withdrawn probably because the postulated therapeutic success was too vague and not permanent.

A Cochrane meta-analysis [53] assessed and valuated 7 trials on the treatment of chronic tinnitus with anticonvulsants encompassing a total of 453 patients. The studies evaluated gabapentin, carbamazepine, lamotrigine, and flunarizine. None of the studies showed a significantly positive effect for one of the pharmaceutics. One trial revealed a significantly negative effect of gabapentin. One trial showed a – non-significant – positive effect of carbamazepine compared to placebo, another one showed an – also non-significant – negative effect. The same is true for flunarizine, positive and negative effects were revealed but without significance. Side effects were described for all tested substances in 18% of the patients. This Cochrane meta-analysis clearly confirms that anticonvulsants are not only ineffective but sometimes even deteriorate the tinnitus-induced distress.

This study should have set an end to the discussion of blood flow in the context of chronic tinnitus, since a clearly higher blood flow of the peripheral vessels can be achieved by vardenfil which is also shown by the side effects. The study is methodologically sound and conducted in a controlled way, however, it is surprising that soft parameters such as effects on sexual life or partnership are not mentioned in the original article. Probably they are responsible for the originally positive statements of the patients that were the reason for the trial.

A high-quality pilot study was conducted at the Charité in Berlin as a double-blind, prospective, randomized, and placebo-controlled trial on the efficacy of a PDE5 inhibitor in tinnitus patients [52] . 42 patients with chronic tinnitus received either 10 mg of vardenafil twice per day for 12 weeks or placebo tablets. The therapeutic success was assessed by means of the tinnitus questionnaire. Neither the questionnaire total score nor the single subgroups showed a significant improvement of the vardenafil group compared to placebo. The authors conclude that the vasodilative effect and the cGMP increase induced by the agent have no impact on the tinnitus symptoms even if hypoxia and oxidative stress play a role in the genesis of tinnitus. Severe side effects were not described, but prolonged erection and swellings of the nasal mucosa were observed. Furthermore, headaches, vertigo, and facial flushing occurred. An influence of the hearing capacity was not observed, neither positive nor negative.

Piribedil, a dopamine agonist, does not significantly improve tinnitus. In a double-blind, placebo-controlled, prospective cross-over study conducted by a team from Regensburg [51] randomized 100 patients with chronic tinnitus to undergo therapy with 50 mg of piribedil and placebo for 90 days each. Piribedil is a dopamine agonist and is applied in Parkinson’s disease. 56 patients finalized the study, the therapeutic success was assessed via the tinnitus handicap inventory (THI) and visual analog scales. In comparison to placebo, no significant improvement could be achieved with piribedil therapy. However, the electro-cochleographic findings and also DPOAE were conspicuous in the piribedil group and different from those of the placebo group: some patients developed a double peak in the electro-cochleography in the CAP, they also had better responses to the therapy, which it was not significant. A suppression of the DPOAE observed in some patients could not be further correlated.

In the following paragraphs, single studies will be in the focus while the few placebo-controlled studies will be reported at the beginning. Afterward evaluations on different substances and substance groups will be presented.

The above-mentioned statements on acute tinnitus are all the more true for pharmaceutical treatment attempts of chronic tinnitus – currently there are no serious options of pharmacotherapy. Although the legislator requires far-reaching and valid studies for the approval of pharmaceutics for certain indications not only in Germany, a high number of drugs are evaluated with regard to their efficacy in chronic tinnitus – however those are nearly exclusively comparative studies. Hereby several or two drugs are compared in terms of their efficacy, only rarely placebo-control is performed. Especially studies that are promoted by pharmaceutical companies are reluctant regarding a comparison to placebo. This aspect was obvious in studies conducted during the last years on the efficacy of the massively promoted ginkgo extract.

The authors of this article do not explain why all patients received benzodiazepine that is considered as causing depression itself beside a high potential of addiction. Such a therapy regime cannot be recommended; even the trial is useless with regard to such a drug cocktail. The fact that it could actually be published (after all it was even The Laryngoscope) does not stand for thorough work of the reviewers!

107 patients with sudden hearing loss and tinnitus were randomly assigned to 3 groups: group 1 received only alprazolam, a benzodiazepine with middle duration of action. Group 2 also received alprazolam and 4 intratympanic cortisone injections, and group 3 additionally received 4 injections of lipo-prostaglandin E. Group 2 (benzodiazepine + intratympanic injection) had the best results, it improved of 75%. In 25.8% of the cases, the tinnitus completely disappeared, the same was observed in 20% of group 3, but only in 9.8% of group 1 [50] .

Both trials are methodologically well performed, but they both do not consider sufficiently the hearing loss and its improvement under therapy. With intratympanic AM-101 treatment, the hearing threshold improved significantly but this fact was not correlated with the improvement of the tinnitus. The efficacy of this NMDA antagonist regarding acute tinnitus is not clear because it is not very convincing if primarily no significant differences are observed, but in a post-hoc analysis groups are found where an effect can be confirmed and then a general efficacy is concluded. In general, tinnitus after middle ear surgery or inflammation does not frequently occur, rather after noise-induced trauma. But in this context, cortisone therapy is also effective. According to this study with intratympanic application of the NMDA antagonist, there was no improvement for the patients because the tinnitus severity did not change, only its loudness improved. This parameter, however, is completely individual and also scientifically only a difficult measure that does not correlate with tinnitus severity. It is astonishing that many authors tend to include a positive recommendation in their abstracts even if actually no effect of the therapy could be confirmed.

In another study performed afterwards in 2015 on this NMDA antagonist, the best dosage was found [49] : in 16 centers in the USA and Europe, 85 patients who had tinnitus for not longer than 3 months after noise-induced hearing loss, barotrauma, acute middle ear infection, or middle ear surgery, underwent intratympanic injection of the drug or placebo in the context of a phase II study conducted in a double-blind, randomized, and placebo-controlled way. Half of the patients received one treatment, the others had 3 injections in weekly intervals (placebo or AM-101). In summary, a low effect size was measured regarding the improvement of tinnitus loudness, more clearly for the group with 3 injections. The follow-up time was 90 days. Regarding the subjectively perceived tinnitus severity measured in an analog scale and the severity measured by means of the tinnitus questionnaire, the effect was much lower. Undesired side effects were mainly local reactions, however, in 17% the tinnitus even deteriorated and 12% had a hearing reduction that occurred due to the intervention and was regressive afterwards. The authors conclude that this treatment is appropriate, probably and in contrast to the previous study as acute treatment also for other tinnitus origins.

Consequently, esketamine hydrochloride (AM-101) was further evaluated in larger multicenter trial for treatment of acute tinnitus [48] . This study was double-blind, prospective, and placebo-controlled, 248 patients between 16 and 65 years were treated with 3 intratympanic injections of AM-101 (high- or low-dose) or placebo on 3 consecutive days. No significant improvement was observed, only patients with tinnitus after noise-induced hearing loss and after middle ear infection experienced improvement with the drug compared to placebo. During the follow-up time of 90 days, the subjectively perceived tinnitus loudness was better with AM-101.

The basis for similar pilot studies are research protocols according to which NMDA receptors are up-regulated in stress situations of the inner ear and cause increased activity of the neural fibers, possibly also tinnitus. In a pilot study [47] , conducted in a multicenter, double-blind, randomized, and placebo-controlled way, the NMDA antagonist esketamine hydrochloride (AM-101) was injected intratympanically in different dosages in 24 patients with acute tinnitus persisting not longer than 3 months. The tinnitus severity was measured by means of the tinnitus questionnaire TQ-12 and assessed 60 days after therapy for the last time. None of the dosages and neither the placebo could improve the tinnitus severity. The self-rated tinnitus loudness and the MML (minimal masking level) changed moderately, also in the placebo group. The authors consider this therapy as a good and safe therapeutic option with tendency to improvement of the tinnitus. The fact that the tinnitus questionnaire as only evaluated measuring instrument did not reveal any improvement, is explained by the aspect that some patients had bilateral tinnitus but only one side was treated so that the distress persisted.

In another study from Korea, 139 patients with acute tinnitus were treated with intratympanic injections of dexamethasone on 4 consecutive days. The patients had no noise-induced hearing loss and no clear (>30 dB) acutely occurring hearing loss. 43 patients (37.7%) lost their tinnitus, 42 had improved results, and in 29 patients the tinnitus remained constant. The authors concluded a good prognosis after intratympanic dexamethasone (ITD) injection [45] , however, no placebo-control was performed so that the significance of this study is low.

70 adult patients with acute tinnitus underwent randomly either intratympanic methylprednisolone or saline solution injection. The treatment was applied once per week for three weeks. Both groups had no pretherapeutic differences regarding age, sex, or tinnitus severity and loudness. After treatment, a significant reduction of tinnitus loudness was observed in both groups. The tinnitus severity, however, did not change significantly. In both groups, pains were observed as side effect, additionally a burning sensation and bitter taste were seen in the prednisolone group. The authors could not confirm a therapeutic benefit, unfortunately, no statement was found on the origin of possibly concomitant hearing loss.

In accordance with the guideline [31] , the therapy of acute tinnitus follows the treatment of acute sudden hearing loss. The same problems are present in both diseases: evidence-based therapy regimes are rarely found, actually the generally concomitant acute hearing loss is treated (with high-dose cortisone, systemic or intratympanic application). If tinnitus occurs acutely without hearing loss, the standard therapy is not recommended because psychosomatic factors such as overstimulation certainly play a decisive role [43] . Even more recent therapeutic approaches such as the intratympanic treatment have no effect on tinnitus as described in a study from Istanbul [44] :

Until a few years ago, medication stimulating the blood flow was considered as standard of tinnitus treatment – not only in the acute stage. Agents such as pentoxifylline even have (until now) an approval for the treatment of hearing disorders and sudden hearing loss; a high-dose ginkgo extract (Tebonin ® 120 mg for ear noises) is approved for adjuvant therapy of tinnitus of vascular and involutive origin [42] – whatever this may be. For both agents, however, there are no scientific proofs for efficacy, nonetheless both substances are still frequently prescribed in cases of chronic tinnitus but not paid by the statutory health insurances. At the same time, numerous other drugs are tested and compared with one another in trials, rarely with placebo; often – mostly sponsored by pharmaceutical companies – recommendations are given that do not withstand scientific evaluation or even meta-analysis.

Since pharmacological solutions cannot be rapidly found in the near future, the basic conditions for effective tinnitus therapy are oriented rather on cochlear and central mechanisms. In his review article, Noreña from Marseille, France, examines all possible origins of tinnitus. They are found on the cochlear level, mainly in the external hair cells, and caused for example by noise trauma or stimulation peaks in the inner hair cells with activation of the NMDA receptors. In the central auditory system, tinnitus develops by aberrant activities that are mostly generated by plastic alterations caused by hearing loss, hyperpolarization of neurons of the thalamus, or a general increase of the central activity of the auditory system with increased influence of even non-auditory irritation [8]. In terms of therapy, this knowledge led to the development of different therapeutic approaches especially in the last years attempting to directly influence the central stimulation patterns or re-organization in the primary or secondary auditory cortex. A recent review article that was presented during the international tinnitus seminar in 2014 in Berlin, describes cortical magnetic stimulations, direct or indirect electric stimulations, but also attempts to influence acoustically those aberrant stimulation patterns [43]. The most important therapeutic approaches will be presented in the following paragraphs and assessed critically with regard to their evidence.

4.1 Inner ear (electro-stimulation – local anesthetics – neural therapy – laser)

In the past, there were many attempts made to directly reach and influence the inner ear. Because of lacking success, those ideas have not been further pursued.

4.1.1 Cochlear electro-stimulation

Some groups examined the effectiveness of electric stimulation of the inner ear. 120 patients with tinnitus and hearing loss were divided into 2 groups in the context of a double-blind and placebo-controlled trial. 80 patients were stimulated via the auditory canal (non-invasive) with frequencies of 250–8,000 Hz for 4 minutes each and an intensity of 1.15 mA, 40 patients were placebo-treated, i.e. without electric current. This trial from Poland revealed that the tinnitus disappeared immediately after treatment in 33.6% of the treated patients and in 6.1% of the placebo group. However, the follow-up showed results that were less favorable but the improvement of the group of electro-stimulation was significant. The changes were assessed by means of a self-designed questionnaire encompassing 20 items [77]. In the context of a trial from Israel, 10 tinnitus patients were stimulated with an intratympanic needle electrode (comparable to the promontory test) and pulses of 100 and 1,800 Hz with amperages of 0–1 mA. The therapy was applied for 30 minutes each on 3 subsequent days. The tinnitus severity improved in 5 patients, but turned to the original status 4 weeks after treatment. Also the audiological tests could not reveal any changes – the authors postulate that this therapy might be useful for some patients; further it has no side effects [78]. The authors consider as effective a suppression of the tinnitus by the (peripheral) stimulation but not masking. This effect can only be temporary. Anyway, this non-randomized and in particular not placebo-controlled study is only an approach without any evidence.

4.1.2 Local anesthetics

Direct injection of local anesthetics or diffusion into the middle ear by means of iontophoresis had no influence on the tinnitus, not even temporarily. Neural therapies with targeted injections “around the ear” can possible relieve contractions or even pains, however, the tinnitus cannot be influenced, neither in terms of loudness nor regarding its concrete bothersome distress. Current and especially valid studies were not published. Success rates published many years ago [79] could never be validated afterwards.

4.1.3 Soft laser – no therapeutic effect on tinnitus

Only the approach that has been propagated for more than 30 years to irradiate the inner ear with a soft laser, i.e. non-cutting bundled light, and to thus influence the tinnitus, still has supporters, and especially sellers. Theoretically, the laser treatment is expected to stimulate energetically the inner ear, some laser therapists even claim to be able to treat hearing impairment [80]. At the end of the last century, some trials proved that each form of laser radiation (different wave lengths, with and without additional ginkgo) cannot effectively influence the tinnitus [81], [82], [83], [84]. Nonetheless, this therapy is still propagated in single cases and even an expensive self-treatment device is marketed (“Tinnitool”). Recent studies, however, confirm again the ineffectiveness. In an analysis from Milan [85] 60 tinnitus patients assigned to two groups were treated either with active laser light (650 nm, 5 mW) or with a dummy in a prospective, randomized, and double-blind trial. All patients had concomitant hearing loss. The therapy was performed for 20 minutes per day over 3 months. The therapy success was assessed by means of the THI, a statistically significant difference could not be revealed between placebo and active laser irradiation. Also tinnitus masking and hearing capacity did not change, however, the tinnitus loudness measured in SL was slightly lower in the laser group. The study is characterized by the fact that also the impaired hearing ability was measured – in all patients, beside other audiometric data. A therapeutic effect cannot be achieved by soft laser therapy, neither with regard to the tinnitus nor as hearing improvement. In Iowa, a double-blind, placebo-controlled trial was conducted that evaluated the effect of laser therapy on hearing functions [86]. 30 patients were randomly assigned to 3 groups and either treated with low level laser (n=9), with placebo (n=10), or not at all (n=10). Laser and placebo treatment were applied in 7 steps to different points of the head for a total of 4 minutes. Before and after treatment, subjective audiometric data were assessed and otoacoustic emission were measured. In none of the groups, altered hearing functions could be revealed neither with nor without treatment. In Europe and especially in Germany, low level laser therapy was heavily propagated about 30 years ago without being able to provide a proof of success. In the USA, it appeared in the last years, rather promoted for improvement of speech understanding. However, success could not scientifically be confirmed even if the presented study only treated very small numbers of cases and furthermore different regions of the head were “irradiated”. In a double-blind, placebo-controlled trial from Austria, 48 patients showed no significant therapeutic effect in comparison to placebo [87]. In 2014, 43 patients were treated in the context of a double-blind, placebo-controlled prospective, randomized trial in Malaysia. The laser device (“Tinnitool”) was actively applied in 22 patients, in 21 patients it was not turned on. Between both groups there were not significant differences. In the active group, 41% observed an improvement, in the placebo group even 59% of the patients [88]. In cases of noise-induced tinnitus, a study from Persia revealed an improvement for the first three months after 20 session with low power laser therapy, afterwards no effect could be found. There was no control group [89]. Laser therapy with a completely ineffective low power laser stimulator is in the market since more than 30 years. Success had never been proven, again and again there are attempts to establish this kind of devices and to sell them at high costs (“Tinnitool” costs more than 1000 Euro!). Meanwhile those commercial business models reach developing countries, however, also there they scientifically prove to be ineffective.

4.2 Cortical interventions – transcranial magnetic stimulation

From the discipline of psychiatric treatment of depression, the idea was taken to influence tinnitus by transcranial magnetic stimulation. This therapy option for chronic tinnitus was evaluated more intensively during the last years, especially different stimulation frequencies and locations were tested. The (psychiatry) group from Regensburg [90] initially assessed the effect of neuronavigated repetitive transcranial magnetic stimulation in chronic tinnitus on the tinnitus-induced distress while the cortical stimulation zones were determined in a methodically complex way and then stimulated specifically at low frequencies. 12 tinnitus patients with increased metabolic activation in Heschl’s convolutions of the cortex underwent radiation with 2,000 stimuli per day and a frequency of 1 Hz; the tinnitus severity could be significantly reduced. No significant improvement could be observed after placebo radiation with a so-called sham coil. Another trial of this group treated 32 patients with repetitive magnetic stimulation either in a low-frequent temporal or high-frequent prefrontal and low-frequent temporal way. Directly after therapy, both groups showed improvements but 3 months later the group with the combined therapy had better results [91]. In contrast, Lee et al. from the USA reported about 8 patients who received rTMS on 5 subsequent days on the left temporoparietal side. A therapeutic effect could not be achieved [92] . In a study from Washington [93], 14 adults (42–59 years) who suffered from tinnitus persisting for more than 6 months were treated with low-frequent (1 Hz) repetitive transcranial magnetic stimulation or a sham probe. The tinnitus severity was assessed by means of the THI. Real treatment improved the value of the THI by 5 points, placebo even by 6 points. Thus the treatment was not more effective than with placebo. Also the group of Tübingen around Plewnia considered the success of rTMS therapy in a rather sober way: 48 patients were treated for 4 weeks, all of them improved a bit based on the tinnitus questionnaire which was observed after placebo as well as after active treatment [94]. On this topic, a current study from Utrecht was published with the conclusion that rTMS was ineffective for treating chronic tinnitus [95]. The randomized, double-blind, placebo-controlled design of the study enrolled 50 patients treated with rTMS (low frequency, 1 Hz or placebo). The evaluation was performed with the tinnitus questionnaire and the THI and VAS. At no time of the follow-up period a significantly better effect of real rTMS could be observed compared to placebo. A recent study from Portland randomly treated 70 tinnitus patients from 2011 to 2014 on 10 subsequent days with active rTMS or placebo [96]. Side effects were not observed, 56% of the actively treated patients (n=32), but also 22% or the patients treated with placebo (n=32) had improved findings regarding tinnitus severity. The therapeutic success was stable during the 26 weeks follow-up.

4.2.1 Magnetic stimulation and location of the coil

Regarding transcranial magnetic stimulation, a strong magnetic field is induced by electric current circulating in the coil. These magnetic waves penetrate the skull and cause an intracranial current reversal. The neuronal activity shall be manipulated by different stimulation variations. Many studies, however, differ regarding the location of the coil. It is placed on areas that are either marked by certain EEG alterations or pre-defined in the functional MRI scan, other studies stimulate always the same regions and vary in terms of stimulus intensity or alternate the stimuli. A Chinese investigation of 22 patients (12 active, 10 placebo) determined the laterality of the auditory cortex with MMN measurements (mismatch negativity), 9 of 12 patients reported a reduction of the tinnitus after magnetic stimulation regarding severity and loudness, however, after 1 months it was no longer significant [97]. The abstract of this study promises more than the actual investigation contains. In the abstract a very promising therapy is mentioned, in the article, and only there, a short sentence is found that the effect was only short-lasting! This article was reviewed and published in a journal with high impact factor!

4.2.2 Magnetic stimulation and different types of stimulus

Numerous publications from the group of Antwerp evaluated different stimulus intensities and rates in always very small patient populations. Only 1 Hz stimulation (applied in 11 patients) seemed to be effective with improvements of 20–40% regarding the tinnitus loudness. Long-term effects were not investigated [98]. Every now and again, the group of Regensburg presents results documenting a rather positive effect. 192 patients who received 10 stimulations over the left auditory or frontal cortex and who were compared to a placebo group (sham treatment) could not report significant reduction of the severity measured by the tinnitus questionnaire, however, the total effect in the follow-up was not significantly better than after placebo treatment [99]. Afterwards, the group assessed only those patients from their collective of 235 patients – a total of 50 patients or 21.3% – who reacted positively on the therapy. Even 2 and 4 years after treatment the results were better than before, however, they became worse in the further course [100]. In another analysis of 538 tinnitus patients, the group looked for predictors for successful treatment with magnetic stimulation [101]. Even if both stimulation types showed an improvement measured with the tinnitus questionnaire, the effect sizes were very low. Predictors for a successful treatment could not be revealed.

4.2.3 Combined application of rTMS and antidepressant

In a study from Turkey, 5 groups containing 15 patients each were compared [102]. Beside a placebo group, 2 groups received transcranial magnetic stimulation with different pulses, 1 group received an additional antidepressant (SSRI), 1 group received only magnetic stimulation. In the follow-up time after six months only patients had an improved tinnitus severity (THI and TSI questionnaire) who had received the antidepressant, with or without rTMS. The data, however, were retrieved from only a very small group of patients which limits the significance – the combination of antidepressant and TMS does not seem to have a generally superior therapeutic success.

4.2.4 Review articles on rTMS

Meanwhile also some meta-analyses are available for this therapy. Two review articles from China and the USA document careful, rather positive effects of non-invasive magnetic stimulation for chronic tinnitus. The Chinese review article found 5 randomized controlled studies with a total of 160 patients but only short follow-up intervals. The authors conclude that long-term effects cannot be determined reliably [103]. An article from Boston assessed 105 publications with a total of 1,815 patients with special attention to side effects. The authors come to the conclusion that side effects are not exactly mentioned or documented. They are generally considered as mild, but occurred in 16.7% [104]. In a review article on repetitive transcranial magnetic stimulation studies were assessed until 2014 and 15 RCTs were evaluated [105]. Generally, the authors state that the studies cannot be easily compared because the numbers of patients were very small, different types and locations of stimulation were used, the rating of the success was different, and in particular no really adequate placebo stimulation was available. Up to now, there is no placebo procedure that the patient does not recognize as such. Nor any study could organize the blinding of the therapeutic staff. This relativizes the “success rates” and according to the authors they should be interpreted most carefully. Nonetheless, the authors state in their conclusion (and of course in the abstract) that rTMS was a therapeutic option, but with modest effect.

4.2.5 Significant noise exposure in magnetic stimulation

An interesting issue which is known from diagnostic MRI, was dealt with by a group from Lyon. They investigated the noise exposure during magnetic stimulation. Significant sound levels were measured, depending from the stimulation intensity peak levels of 120–140 dB(A) were measured with a middle level of 90–100 dB(A). Further, it was interesting that the coil used in the trials as placebo (sham) was quieter of up to 40% [106]. This article confirms that ear protection should be worn during therapy, and it also tries to reveal a correlation between active stimulation and auditory-cortical reactions on noise exposure. This study raises important questions regarding noise exposure by therapy and the simultaneously occurring cortical reactions on noise exposure alone. At the same time the sham stimulation is even more doubtful because this coil does not become warm and is much quieter – which actually nullifies the placebo effect.

4.2.6 Can rTMS be effective?

An interesting and pioneering article was published by the research group from Groningen on the cortical plasticity with the question if hyperactivity of the auditory cortex occurs only in tinnitus patients or if it is a general cortical property. In the studies on transcranial magnetic stimulation, this hyperactivity is valued as correlate of the tinnitus. The cortical metabolism was measured by means of FDG-PET in 20 tinnitus patients and compared with 19 persons without tinnitus. In both groups, the metabolic activity in the left auditory cortex was higher than in the primary right one, for the secondary auditory cortex, the metabolic activity was reciprocal, on the right side higher than on the left. However, there was no difference between tinnitus patients and control persons; thus it seems that they will have to be considered as regular reaction of a healthy brain [107]. But the effects of rTMS on the cerebral function do not seem to be clear. The group from Regensburg analyzed the literature with regard to possible alterations of the motor cortex after different magnetic stimulations, but they did not find systematic and comprehensible alterations of the cortical irritability [108]. In a guideline from several European countries, the statement on repetitive transcranial magnetic stimulation is given that an analgesic effect on the primary motor cortex is evident (level A). For depressions, a level B effect could be assumed with probable effectiveness, but for tinnitus treatment only a level C effect could be observed (probably effective) [109]. In summary, magnetic stimulation (rTMS) is certainly a scientifically sound approach that is very complex and has its weaknesses with regard to studies since the patients can recognize if a coil is active or not (missing warmth and less loud) which makes placebo stimulation not real. For all therapy studies, higher case numbers and especially longer follow-up times must be required in order to really identify and prove long-lasting and effective therapeutic approaches for chronic tinnitus. However, the results of rTMS are described in contradictory ways. While the group from Regensburg reports about therapeutic success with different application types and series for several years, data from the USA do not confirm these findings with still very few case numbers. But it is surprising that none of the numerous studies in this field includes audiological examinations. If magnetic stimulation changes cortical (or subcortical) reactions, measurable changes should be found for example in subjective central hearing tests or cortically evoked potentials and contingent negative variations. So only the subjectively perceived severity is the criterion of success. In conclusion, it can be said that transcranial magnetic stimulation is a promising therapeutic option for many study groups, especially psychiatrist. However, it is effective only for short intervals and not superior to placebo. Often secondary result parameters are used in those studies (especially in trials from Regensburg and Antwerp) that reveal better success for the stimulation group. If the results are not sufficient, they are “corrected” (calculated in another way). In the past, this type of therapy was tentatively applied in psychiatric centers for treating depressions, in comparable studies with similarly poor success. In summary, the very contradictory study results of the last years seem to show that rTMS for tinnitus is effective for the time of treatment but long-term efficacy cannot be confirmed. This statement is in accordance to the fact that the original hypothesis is not true that hyperactivities of the auditory cortex were pathognomonic for tinnitus. However, many treatment attempts are based hereupon with influencing the cortex by magnetic stimulation as well as directly or indirectly with noises (“neurostimulation”). Even if a recently developed European guideline confirms a possible (level C) effect, there are meanwhile sound studies like the one from the Netherlands [107] proving that transcranial magnetic stimulation has no effect on chronic tinnitus.

4.3 Electro-stimulation

Another possibility of directly influencing cortical structures consists of applying electrical stimuli. This aims at influencing aberrant electrical activity in the cortex. Assumed synchronicity could get out of the rhythm or the electrical stimulation could initialize counteracting stimuli that might then eliminate or influence the reaction of tinnitus. This stimulation can be applied transcutaneously or directly into the cortex, but previously it has to be clarified which region shall be stimulated. These therapeutic approaches have been used for many diseases in psychiatrics – all of them with rather poor success. Also for tinnitus treatment we look back to multiple investigations. Based on the knowledge that cortical alterations are at least responsible for the persistence of tinnitus, the attempt is made in many centers to stimulate the auditory cortex or other cerebral regions directly or transcranially by electric current.

4.3.1 Transcutaneous, transcranial electro-stimulation

The efficacy of transcutaneous electro-stimulation for tinnitus treatment was evaluated in a small patient population (n=31) with placebo control. Low-frequent alternating current (<100 Hz) with intensities of 50–2,000 mA and a pulse rate of 30 Hz was applied. Significant improvement in comparison to the placebo group could not be achieved [110]. In a trial from Regensburg [111], 32 patients with chronic therapy-refractory tinnitus underwent bifrontal electro-stimulation with 1.5 mA for 30 minutes 6 times, twice per week each. The results were assessed by means of the tinnitus questionnaire. There was no improvement of the tinnitus severity, not even of the depression. Loudness and annoyance improved minimally based on subjective impression. In a double-blind and placebo-controlled study from Belgium [112], 20 patients with chronic tinnitus received transcranial electro-stimulation for 20 minutes either anodically, cathodically, or with sham stimulation (placebo). During treatment, the intensity of the tinnitus changed, the severity was not assessed. Especially the anodic stimulation reduced the intensity of the tinnitus during treatment, some patients even reported an effect of several days. The study group from Antwerp presented a study of 111 patients with tinnitus persisting for more than 1 year; 36 patients received transcranial direct current stimulation (tDCS), 37 had transcranial alternating current stimulation (tACS), and 38 received transcranial random noise stimulation (tRNS) (with randomly stimulating and changed alternating current voltage). Significant improvement (VAS tinnitus loudness and severity) was observed especially in the context of random noise stimulation (tRNS) [113]. These types of treatment and electric stimulations have certainly less side effects; the evaluation revealed some improvements immediately after stimulation, but longer lasting success could not be achieved or it was not even assessed. Furthermore, VAS and non-validated questionnaires are a weak instrument for evaluation. A group from Switzerland investigated the effect of transcranial direct current stimulation in a double-blind and placebo-controlled study with 42 patients [114]. The cathode was placed on the auditory cortex, the anode in the prefrontal region. No side effects were observed, but there were neither effects on the tinnitus. In 21 out of 26 patients (77.8%) from a New Zealand study, an improvement of the tinnitus loudness and severity of at least 1 point after 2 sessions of tDCS (different locations and intensities) was achieved [115]. Japanese researchers investigated the influence of the connectivity by tDCS [116]. Since the connection between the left and the right auditory cortex seems to be less developed in tinnitus patients in comparison to regularly hearing patients, 9 tinnitus patients were compared to 9 control persons. After tDCS, the connectivity between auditory cortex and somato-sensory and motor brain areas was reduced in tinnitus patients, while this connection remained strong in the control group. It is unknown why this study did not provide data on the hearing loss of the tinnitus patients so that no statement could be given on the correlation of hearing loss and tinnitus.

4.3.2 Invasive cortical tinnitus therapy

Many tinnitus patients are ready to undergo even invasive procedures in order to get rid of their tinnitus. In an investigation from California, 800 firefighters were asked if they would be ready to undergo intracranial surgical radiation with the so-called gamma knife at the caudate nucleus to treat their tinnitus. The readiness depended on the prognostic success: in case of 100% success, 60% of the interviewed persons would agree to this therapy; in case of 75% success, there were still 43%. Below those values, the readiness was significantly lower [117].

4.3.3 Intracortical implantation of electrodes

While transcutaneous electro-stimulation has no effect on the tinnitus perception according to the available publications, the question remains if direct intracortical stimulation is more promising. At this point, an article from 2011 will be cited as example of very far-reaching therapeutic options. In Antwerp, electrodes were implanted in the secondary auditory cortex in 43 tinnitus patients who had previously responded positively on transcranial magnetic stimulation (rTMS) for a certain time [118]. Three days after implantation, the electrodes were stimulated with 40 Hz bursts. If no response was observed, the burst frequency was changed until a reaction on the tinnitus was perceived. The success was measured by means of a visual analog scale (VAS) before and after treatment, there was a placebo group but it could not be evaluated because of reasons that were not explained in the article. A total of 29 patients out of 43 responded to electro-stimulation, but only 8 reported a real improvement. Statistically, the VAS data revealed 67% success rate, 33% of the patients did not respond with regard to their tinnitus. In terms of side effects, 3 patients developed epileptic seizures, 1 patient had brain hemorrhage and consecutive difficulties with speaking and finding words (however, the tinnitus had improved), 1 patient developed a brain abscess that had to be treated surgically (and the tinnitus became worse). Fortunately, such therapeutic adventures are very rare and probably cannot be carried out in our country because there are efficient and responsible ethical committees. Already the assessment of such invasive procedures only with VAS shows that a real evaluation is not performed. It remains unclear why the placebo group could not be evaluated and how it was stimulated. Finally the side effects and complications are mentioned but the authors still think that a good and potentially promising therapy is found. The group from Antwerp used transcranial magnetic stimulation for exact definition of the location of implanted electrodes for tinnitus suppression (1 patient) [119]. A direct electro-stimulation of the auditory cortex led to significant reduction of the tinnitus in this patient who underwent implantation of an intracortical electrode. Another patient with extracortical electro-stimulation experienced a short-term, non-persisting tinnitus reduction. The implantation of electrodes into the defined region was carried out after intensive diagnostics with functional MRI examination of the tonotopic tinnitus representation in the cortex [120]. Another study with 8 patients who suffered from chronic tinnitus and who underwent neurosurgical electrode implantation after fMRI examination and electro-stimulation for 2 weeks, observed similar results of an at least partial tinnitus reduction. In this study, the stimulation was interrupted after 2 weeks and replaced by sham stimulation. In this period, no further improvement was found [121]. In a double-blind, placebo-controlled, and randomized cross-over study from Bordeaux, 9 patients with severe unilateral tinnitus underwent electrode implantation in the auditory cortex under general anesthesia [122]. This electrode was connected to a stimulator implanted in the pectoralis region. For 4 months, the patients were biphasically stimulated, afterwards there were randomly assigned to 2 groups and either placebo or really stimulated. After a washout they were again stimulated cross-over, therapeutic success was measured by means of a questionnaire (structured tinnitus interview – STI). One patients had to be explanted because of severe psychic decompensation; 3 patients were explanted after the end of the study, whereas 5 were stimulated for further 3 years. During the open phase, the tinnitus improved in 5 patients, 2 patients reported about deterioration of the findings. In the following control phase, also improvement was found, however, it was seen in the placebo group as well as in the directly stimulated group. Side effects were not observed, in particular no changes in hearing. The authors conclude from this investigation that direct electro-stimulation of the auditory cortex is associated with significant placebo effect, already because of the surgery. A therapeutic effect on tinnitus severity is not achieved in consideration of this placebo effect. In summary, regardless the side effects, all those studies are not appropriate to confirm a valid and significant therapeutic success because of the low numbers of patients (n=2 or n=8 or 9), also because the placebo effect must not be neglected. Since electro-stimulation is perceived by the patient, even switching off the electrode is no real placebo. In addition, many articles do not mention possible risks and side effects.

4.3.4 “Deep brain stimulation”

A recent review article from Maastricht shows possible therapeutic perspectives by deep brain stimulation for which an electrical pulse generator is implanted into cerebral structures (“brain pacemaker”). This therapy had positive effects in the symptomatic treatment of therapy refractory Parkinson’s disease and shall be evaluated for application in tinnitus treatment. It is based on the reflection that the stimulation zones in tinnitus patients are found in very different brain areas so that they would have to be stimulated individually [123]. Investigations on this topic are currently not available. Anatomically and surgically, such a targeted implantation of electrodes seems to be possible, if it turned out to be therapeutically reasonable in cases of tinnitus [124]. The German practice guideline on deep brain stimulation, however, does not include the indication of tinnitus [125]. Deep brain stimulation in humans has not been validly investigated with regard to tinnitus, because of its invasiveness this procedure has to be intensively and carefully discussed. In contrast, a well-designed study from Bordeaux [122] reveals that the invasive procedure of direct intracortical stimulation temporarily reduced the tinnitus severity in some patients, but that it is not better than placebo stimulation. It is surprising that no side effect occurred since other neurosurgeons observed severe complications after electrode implantation [118]. In the context of transcranial stimulation, which is significantly less invasive, no relevant effects with regard to tinnitus could be found in this methodically good investigation performed in 42 patients. Studies with an improvement of the severity by 1 point (!) and doubtful changes of non-confirmed connectivity are simply irrelevant. Furthermore, no data is retrieved in these studies on the hearing loss which would be essential for the evaluation of the connectivity. Again the question must be asked how so poor studies pass the review process and are published in reputed journals.

4.4 Vagus nerve stimulation

Transcutaneous vagus nerve stimulation is considered as new and quasi most recent type of tinnitus treatment. Medially highly appreciated and published in the journal Nature, those studies referred only to animal experiments until now. These animal experiments led to the knowledge that animals learn more easily when the vagus nerve is electrically stimulated, simultaneously to the exercises [126]. For the treatment of chronic tinnitus in humans, the ear noise shall be modified by sound therapy that is accompanied by direct vagus nerve stimulation. In a pilot study on the general feasibility, 24 tinnitus patients underwent vagus nerve stimulation for 3–10 weeks and were monitored regarding cardiac complications in order to explore clearance and safety of this therapy. Two severe cardiac complications occurred, the therapy seemed to shorten the QRS complex in the EEG, in primarily healthy persons, arrhythmia did not occur [127]. Another feasibility study was based on the question if such a treatment was really possible in humans. In New Zealand, 10 tinnitus patients were selected and electrodes were implanted at the left vagus nerve at the neck. The patients heard noises for 20 days for 2.5 hours per day that did not correspond to the tinnitus frequency. At the same time the vagus nerve was electrically stimulated. The therapy was well tolerated. Side effects did not occur. According to the authors, this means that this therapy is safe and feasible [128]. In general, a stimulation of the vagus nerve leads to a reduction of the sympathetic innervation which was proven by an article from Leeds evaluating 48 healthy test persons [129]. A multicenter study on vagus nerve stimulation that was initiated in the USA 2 years ago is currently not finished or even published. The homepage of the company does not reveal recent statements even if they are expected to be present. So only a case study from New Zealand was presented until now (and published in a highly ranked journal) [130]. In a 59-year-old patient who suffered extremely from his tinnitus and who had undergone – unsuccessful – intracortical implantation of electrodes, the vagus nerve stimulator was implanted and he received sound therapy for 4 weeks. The achieved improvement (measured by THI) and also the improved depression persisted for 2 months, afterwards the severity of the findings was the same as before therapy. Of course, the authors postulate that the vagus nerve stimulation could become an e