Great
article,
@Hazel. It still holds its relevance 4 years later as it really delves into the complexities of tinnitus research and the extreme challenges hindering the discovery of a cure still to this day.
Suppose I would like to go over some of the more interesting bits from the article and offer some my personal thoughts too. It would be nice to hear everyone else's perspective on this matter as well which is the reason for bumping this thread.
Ambiguous Tinnitus Definitions and Subtyping
Multiple definitions of tinnitus have been published from "ringing or buzzing in the ears" (Oxford Dictionary) to "the conscious experience of a sound that originates in the head of its owner" (McFadden, 1982) or "the conscious perception of an auditory sensation in the absence of a corresponding external stimulus" (Baguley et al., 2013). None of these definitions are entirely fit for purpose. Ringing in the ears is clearly too simplistic – awareness of tinnitus does not have to be within the ears and many sounds other than ringing are reported. The other definitions (McFadden, 1982; Baguley et al., 2013) are more accurate descriptors of tinnitus, but would include the auditory hallucinations seen in some forms of psychiatric illness. Also, some examples of pulsatile tinnitus are generated mechanically, for example, by muscular or vascular activity. Similarly, some examples of low frequency noise complaint are responses to genuine low-frequency noise in the person's environment though others are probably phantom perceptions which would fall underneath the tinnitus umbrella (Baguley et al., 2016).
The various subdivisions of negative reaction to both real and phantom sounds are depicted graphically in Figure 3.
Most tinnitus trials are conducted in subjects with persistent spontaneous tinnitus also known as subjective idiopathic tinnitus. This is the group of tinnitus sufferers whose tinnitus is non-pulsatile and not related to a small number of specific medical conditions or syndromes including Meniere's disease, otosclerosis and vestibular schwannoma. However, this group is extremely unlikely to be a homogeneous population either in terms of their tinnitus pathogenesis or tinnitus experience. Describing tinnitus as idiopathic in particular seems inappropriate: the majority of subjects presenting with tinnitus have a hearing loss measurable with conventional pure tone audiometry (Sanchez et al., 2005; Mazurek et al., 2010). There are some patients, perhaps up to one in 10, who have tinnitus in association with normal pure tone audiometry. However, when more sophisticated investigations of cochlear function such as high frequency audiometry (Vielsmeier et al., 2015) or threshold equalizing in noise (TEN) testing (Weisz et al., 2006) are undertaken, most if not all are found to have defects of peripheral auditory function. It is often taken for granted that one form of sensorineural hearing loss is much the same as another, but is this really correct? Is tinnitus arising in someone with noise induced hearing loss identical to the tinnitus in someone whose sensorineural hearing loss is classified as age-related hearing loss, ototoxic medication induced hearing loss or post-head injury hearing loss? Without fully understanding the pathophysiology of different forms of sensorineural hearing loss and its relationship to tinnitus, researchers may well be undertaking studies on heterogenous patient populations that have different underlying mechanisms. This runs a significant risk that subtle treatment effects for specific groups may be missed in the overall picture. Thus, it is possible that effective treatments for some forms of tinnitus already exist, but this effect has been overlooked because results for multiple subtypes have been analyzed as a single group. An interesting example of tinnitus research where test-subject heterogeneity may have affected trial outcome is the story of gabapentin. As its name suggests, gabapentin was initially thought to be a GABA receptor agonist but is now recognized to have its effect by acting on a subsection of voltage-gated calcium channels. It is marketed as an anti-epileptic drug and is also used in the management of certain types of pain. It has been explored in both human and animal studies for possible use in tinnitus. The animal study (Bauer and Brozoski, 2001) suggested that gabapentin was effective at attenuating tinnitus secondary to noise exposure. A subsequent single blind human study undertaken by the same team (Bauer and Brozoski, 2006) suggested that gabapentin was also effective in humans, particularly those whose tinnitus etiology was associated with acoustic trauma. A double-blind trial also reported that gabapentin was effective in treating tinnitus secondary to acoustic trauma (Goljanian Tabrizi et al., 2017). Several other studies, however, have not found gabapentin to be effective (Piccirillo et al., 2007; Witsell et al., 2007; Dehkordi et al., 2011). Only one of these studies divided their participants into those who had experienced significant noise exposure and those who had not (Dehkordi et al., 2011). A history of noise exposure did not affect outcome in this study, though the number of participants who reported sound exposure was low: 16 reported being in noisy environments and 6 reported exposure to explosions. With this conflicting evidence, a large study with robust etiological subtyping of participants would seem the logical next step.
According to the article, a simplistic definition like 'ringing or buzzing' is problematic because it doesn't capture the full range of sound that tinnitus sufferers can experience while other more accurate definitions tend to include auditory hallucinations and also sounds caused by other various factors like muscle or vascular activity.
I definitely think the absence of a universally agreed upon definition for tinnitus among researchers is a big issue too. This should honestly be the first step before anything else because it will hinder our ability to to accurately identify and categorize the condition properly.
There's also '
tinnitus disorder' which is a term that
@Hazel kept coming across during her most recent visit to The Association for Research in Otolaryngology (ARO) 2023 conference. She brings up many good points as to why that kind of terminology for tinnitus is troublesome in regards to research here. Personally, I think it's important to understand the different types of tinnitus (looking at you reactive tinnitus) and how they relate to hearing loss especially. This way, researchers can better tailor treatments to specific subgroups of tinnitus sufferers which is what the article is emphasizing here.
Tinnitus Measures and Biomarkers
As stated in the previous section, one recurring problem with tinnitus research is that there is no objective way of determining whether someone has tinnitus, no objective way of determining the severity of that tinnitus and no objective way of assessing whether treatments improve tinnitus. A recent systematic review examined the work to date on trying to find suitable objective measures of tinnitus (Jackson et al., 2019). The review identified 21 articles, studying objective tests that included blood tests, electrophysiological measures, radiological measures and balance tests. The review concluded that the quality of evidence was generally poor and had failed to identify any reliable or reproducible objective measures of tinnitus. A biomarker can be defined as "a characteristic that can be objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes or pharmacological responses to a therapeutic intervention" (Puntmann, 2009). Although this may seem to be another way of describing an objective measure of tinnitus, there are distinctions: a suitable biomarker for drug effect or relevant neural process may not necessarily be a measure of tinnitus or tinnitus pathology. Various candidates for a tinnitus biomarker have been considered, including otoacoustic emission testing, auditory brainstem responses (ABR), gap-prepulse inhibition to acoustic startle, pupillometry, functional imaging, magnetoencephalography, genetic markers, blood or saliva components and markers of stress. Nothing has yet been shown to offer the necessary specificity and sensitivity to be used as a biomarker in tinnitus treatment. though some studies have seemed tantalizingly close to discovering a biomarker: ABR studies in tinnitus patients have shown increased latency and reduced amplitude of ABR Wave I compared to measurements in non-tinnitus control subjects (Milloy et al., 2017; Bramhall et al., 2019). However, the findings have shown considerable variability and lack of consistency between studies, suggesting that further work in this area is needed. There have also been some interesting preliminary findings in genetic studies of tinnitus patients. A twins study (Maas et al., 2017) found evidence supporting a degree of heritability in certain forms of tinnitus. A Swedish study (Cederroth et al., 2019) attempted to disentangle the relative contributions of genetic and environmental factors in medically diagnosed tinnitus patients by exploring a large cohort of people who had been adopted as children. This study suggested that clinically significant tinnitus is associated with genetic factors, with a heritability of 32% but that there is no association between shared-environment factors. There are, however, other studies that provide conflicting evidence regarding the genetic contribution to developing tinnitus and this is another research area deserving more detailed exploration.
One factor that hampers work into finding biomarkers is that we do not yet have a large database of the non-audiological phenotypes of tinnitus patients: collecting data such as the biochemical, radiological and genetic characteristics of large numbers of tinnitus patients has not been undertaken. Ideally a biobank dedicated to tinnitus patients should be created (Cederroth et al., 2017; Szczepek et al., 2019).
Whilst work using genetics to identify pharmacological targets is in its infancy (Cook et al., 2014; Vona et al., 2017; Morgan et al., 2018), it is reasonable to expect that further knowledge regarding the genetic contribution to clinically significant tinnitus would be of considerable value.
Without suitable objective markers or biomarkers, tinnitus research in humans currently uses a range of audiometric and self-report questionnaire measures to assess tinnitus severity and treatment effect. Multiple such tools are available and there is no consensus regarding optimum datasets for clinical research. This makes subsequent comparison of trials and meta-analysis of data problematic. A recent multinational working group has tried to address this (Hall et al., 2015, 2019a; Fackrell et al., 2017) and has proposed a basic portfolio of tinnitus "domains" that should constitute a core outcome set for different types of tinnitus research (Hall et al., 2018). Whilst this suggestion is laudable, it remains to be seen if the tinnitus research community adopts these recommendations and it does not provide the unequivocal objective measure that the pharmaceutical industry desires.
A further limitation of the current tools for assessing tinnitus impact is the reliability and repeatability of such measures: self-report measures of tinnitus have an associated risk of variability, as they supply a momentary snapshot whereas the experience of tinnitus changes with time and context. One approach to reducing that is to perform Ecological Momentary Assessment (EMA) (Goldberg et al., 2017; Probst et al., 2017), a technique also used in anxiety, stress, and pain trials (Yang et al., 2019). Evidence regarding the utility of EMA in tinnitus trials is emergent at present.
I and a few others here have been harping on this for a while now, but unless we finally have a surefire way to objectively diagnose and measure the severity of tinnitus, the quest for effective treatments, especially for a significant portion of tinnitus sufferers, will continue to disappoint over and over again. I think it's essential for everyone (especially researchers) to come to terms with this reality.
Relying on these subjective and unreliable self-reporting measures, that are often influenced by the placebo effect, is no longer acceptable... well to me at least. Should be for all you too. It has failed big time in tinnitus research and I'm so done with it. It should only serve as a complementary tool to objective measurements (once we have them) only.
Self-reporting measures for assessing tinnitus treatments has got to be one of the lamest thing I've ever seen in the field of research. You can just see all the issues and limitations that arise when having to solely rely on them.
1. Placebo effect - tinnitus treatments easily falls for placebos for some reason. Likely because people expect said treatment to work, so they may perceive some kind of improvement even if it has no real effect.
2. Lack of objectivity - self-reporting measures lack the neutrality of measurable data. It's the proper way to evaluate treatment outcomes like with any other field of research.
3. Memory bias - if I was a participant in one of these trials, I probably wouldn't recall my tinnitus experience over time. I think there might be a lot of potential biases in reporting treatment effects here. I thought NAC calmed my tinnitus for a bit, but now I'm don't recall how well it did. My tinnitus is still shit. Case in point.
4. Variability - people's tinnitus can change over short periods of time (I'm talking hours, days, and weeks), which makes it really hard to capture the dynamic nature of this condition using infrequent self-reporting questionnaires.
5. Difficulty to compare things - it's extremely hard to compare the effectiveness of different treatments across multiple studies.
I'm sure there's more problems with self reports...
This brings me to my next point: I think instead of searching for treatments right now, the top priority should be on developing objective methods to detect, perceive, and measure tinnitus and its severity levels. How much longer must we all endure the repetition of disappointment with each new tinnitus treatment initiative that comes our way? It's been long enough.
If we don't address this issue, we will never know, with the highest level of confidence whether treatments are truly effective. We need reliable biomarkers and objective measurements ASAP if we want to break this vicious cycle of disappointment and failure. Treatments can wait.
Animal Models of Tinnitus
Animal models have become widely used in tinnitus research, particularly research regarding tinnitus pathogenesis and research into pharmaceutical treatment of tinnitus (von der Behrens, 2014; Eggermont and Roberts, 2015). Yet tinnitus research literature has several instances where apparently effective treatments in animal models have failed to work in humans. Memantine is an antagonist of NMDA glutamate receptors, used in some cases of dementia. Experimental evidence suggested that it is effective in treating tinnitus arising in rats after acoustic trauma (Zheng et al., 2012). A randomized, double-blind study in humans, however, showed no significant change in the primary outcome measure relative to placebo (Figueiredo et al., 2008). Esketamine, the S(+) enantiomer of ketamine, is another NMDA glutamate receptor antagonist that has been explored for use in acute tinnitus, administered as an intratympanic injection. Despite promising animal data (Bing et al., 2015) and initially optimistic human work a large-scale human study, TACTT3-Trial14 failed to show efficacy. AUT00063 is an experimental drug that acts as a modulator of the Kv3.1 subtype of potassium channels. Animal research suggests that the drug is very effective at reducing hyperactivity in the auditory brainstem after noise exposure in rodents (Anderson et al., 2018; Glait et al., 2018) and hence might be expected to be effective against similarly generated tinnitus in humans. However, a randomized controlled trial of AUT00063 in humans with subjective tinnitus, QUIET-1,
was halted because of lack of efficacy (Hall et al., 2019b). This discrepancy between animal models of tinnitus and clinical trials in humans has various possible explanations: firstly, the pathophysiology of tinnitus in humans may be different from that of laboratory animals (and there remain significant questions about whether the animals do experience tinnitus, and whether our methods for detecting the symptom are reasonable). Secondly, where a drug has failed to show efficacy in a human clinical trial after successful animal studies, it is important to be sure that the drug adequately engaged the pharmacological target in humans. The absence of suitable translational biomarkers is a major hurdle to satisfying this requirement. Thirdly, animal studies and human studies measure different things: animal studies generally use either behavioral tests or gap-prepulse inhibition of the acoustic startle reflex (Galazyuk and Hébert, 2015) to define the presence or absence of tinnitus whereas human studies use self-report and quantify the tinnitus using questionnaires, rating scales or psychoacoustic measures such as tinnitus loudness matching. There is currently no translationally valid outcome measure that can be used in both human and animal studies. Fourthly, some animal studies use outcome measures that may not be detecting tinnitus: the QUIET-1 study measured neural hyperactivity in the dorsal cochlear nucleus of hamsters (Glait et al., 2018). Although the authors argue persuasively that this neural hyperactivity is indicative of tinnitus, other explanations are possible, and the finding could represent hyperacusis rather than tinnitus. Finally, animal studies are very often limited to acute dosing with drugs, whereas clinical trials in humans explore efficacy after multiple days or weeks of dosing. It is important to check in animals that the drug effect does not reduce after chronic dosing which might explain why no efficacy is seen in chronic studies in patients.
It is important to observe that there have been studies where animal and human tinnitus research concur, and it would be wrong to dismiss animal research. Examples where there is positive evidence to support the translational value of animal research prior to human trials include bimodal stimulation using either sound and electrical stimulation of the cervical or trigeminal nerves (Marks et al., 2018) or sound and electrical stimulation of the vagus nerve (Engineer et al., 2011; Tyler et al., 2017).
Animal studies have their fair share of issues, which was somewhat eye opening to me when I read it. The use of animal models in tinnitus research doesn't necessarily mean that it will translate well in testing for potential drug treatments for tinnitus. There have instances where treatments that seemed effective in animals failed to work in humans. Drugs like Memantine and Esketamine for instance, were both showed promise in animal studies, but failed to yield any good results in human trials. This begs the question... what reasons for these disparities between animals models and human trials? This is obviously a complex issue and I'm not going pretend like I know it , but I do agree with the article's explanation that it might be due to differences in the underlying causes of tinnitus in humans compared to animals.
But mostly importantly, the way tinnitus is measured in animal studies (behavior tests and reflexes) differs greatly from how it's self reported and quantified in these questionnaires and whatever other rating scales being used for humans. I keep going back to this, but if we seek to make real progress with condition, it's imperative to have objective measurements pronto!
Right Drug, Wrong Time?
It has long been suggested that tinnitus pathogenesis is a two-stage process: an initial ignition which can be anywhere in the auditory system including the cochlea, followed by a secondary process of promotion which occurs in the central auditory system and maintains the prominence of the percept (Baguley, 2006). Inherent in this hypothesis is the suggestion that there may be different therapeutic targets, depending on the stage of the tinnitus. Thus, cases of tinnitus ignited by damage to the peripheral auditory system, may benefit from drugs aimed at the cochlea, given at or soon after onset of the symptom, whereas established tinnitus may need centrally acting drugs. What is not clear, is the time frame for the change from peripheral to central targets. Guitton et al. (2003) demonstrated in a rat model that an NMDA antagonist, gacyclidine, administered to the cochlea prevented salicylate induced tinnitus when given simultaneously. As discussed above, Bing et al. (2015) produced data in an animal model suggesting that an NMDA antagonist might benefit noise induced tinnitus. In this trial, the drug was administered 2 days after noise trauma. Subsequent human trials such as TACTT3 failed to demonstrate efficacy but included subjects who had developed their tinnitus up to 3 months previously. This topic regarding potential optimal therapeutic windows needs further exploration.
I've seen this issue raised many times now and it's still worth acknowledging, the idea that tinnitus development includes two stages, an initial onset issue likely in the ear followed by a secondary process in the central auditory system (that sustains the perception of tinnitus). Basically implying that there might have to be different ways to treat tinnitus depending on which stage you're in. The problem here as mentioned in the article, is that we don't precisely know when the transition from the ear to the central auditory system happens
So I agree with most of what this article is saying. I think addressing these hurdles is extremely vital for making progress in tinnitus research and ultimately finding effective solutions. I think it's time to prioritize clearer definitions, objective measurements, and lastly, customized treatment approaches to break this annoying cycle of disappointment. IMO, it's the only way if we ever want to advance our understanding of this complex condition.