Introduction

Many lines of evidence support the hypothesis that essential tremor (ET) is a progressive disease that is sustained by a neurodegenerative process. The neurodegenerative nature of ET is attributed to a pathologic process involving the cerebellum and/or brainstem (as demonstrated by anatomic and neuroimaging), progression of symptom severity with advancing disease, and the lack of spontaneous remission of this condition.

Pathologic evidence

The term ‘neurodegeneration’ refers to evidence of neuronal cell death. Probably, because of the general belief that ET is a functional disease (despite its high prevalence), very few studies on ET pathology had been performed until 2004, when a series of 20 subjects with a clinical diagnosis of ET, followed up for a period of 32 years, was described (Rajput et al. 2004). Despite the fact that no brain pathology was found in these ET patients, the following studies have demonstrated the presence of structural changes in the cerebellum, due to a reduction of Punkinje cells (PCs) and increases in torpedoes and hairy basket count, as well as structural changes in the brainstem due to an increased deposition of Lewy bodies (LB) in the locus coeruleus (LC) (Axelrad et al. 2008; Erickson-Davis et al. 2010; Louis et al. 2005, 2006a, 2006b, 2007, 2009a, 2011c; Shill et al. 2008). In this respect, a more relevant study by Axelrad et al. (2008) disclosed a seven times increased number of torpedoes in the PCs of ET subjects with respect to age-matched controls and two times with respect to Parkinson’s disease (PD) and Alzheimer’s disease (AD) patients. Recently, an increased number of torpedoes has been shown in the cerebellar vermis of ET brains, mainly in patients with voice, neck, and jaw tremor (Louis et al. 2011c). The torpedoes could be a marker of axonal damage or regenerative process, but this is still controversial; although the significance of torpedoes is not completely understood, converging evidence supports the hypothesis that they are an indirect sign of axonal PC degeneration; however, this sign is not specific to ET, and is present in other clearly neurodegenerative conditions, such as progressive supranuclear palsy, dentatorubropallidoluysian atrophy (Matsumoto et al. 1998), and fragile X associated tremor/ataxia (Greco et al. 2002). Consequently, the number of PCs as a marker of neurodegeneration in ET has been an obvious subject of study. The heterogeneity of populations studied and the difference in methods of analysis make the results obtained in these studies not easily comparable and sometimes contradictory (Louis et al. 2011d; Rajput and Rajput 2011; Rajput et al. 2011). At the Essential Tremor Centralized Brain Repository of New York, the largest series available of ET patient brains (N 33) was compared with 21 controls, with the result that at least two pathologic phenotypes emerged: a cerebellar one (involving 75 % of patients) with reduced PCs and increased torpedoes, characterized by younger age of tremor onset, gait difficulty, and positive family history of ET; and an LB phenotype, with LBs mainly located in the LC (Louis et al. 2007). One year later, a clear reduction in PC linear density in a group of eight ET patients without evidence of LBs compared with 11 controls was confirmed (Axelrad et al. 2008). The reason for the PC reduction in one group of ET patients and the increase in LBs in the LC in another group of ET patients has not yet been clinically defined, but it does confirm the heterogeneity of this disease, which clinicians are already well aware of.

Major criticisms of the above-mentioned anatomopathologic studies are the advanced age of patients evaluated and the occurrence of pathologic aspects that could also be detected in subjects without evidence of neurologic disorders. In response to these criticisms, firstly, most of the listed studies that comprised age-matched normal subjects and compared with them ET patients, showed a decrease of PCs and an increase of torpedoes in a subgroup of patients or an increase of LBs in another subgroup of patients; secondly, the consideration that LBs are present in about 20 % of ET patients but also in elderly controls in a similar percentage does not take into account the anatomic distribution of LBs. As observed by Louis (2010), the majority of LBs in ET patients is located in the LC, while in elderly normal subjects, the LC is only marginally involved.

Neuroimaging

Neuroimaging studies represent a surrogate in vivo biomarker of degeneration in various neurodegenerative disorders, and despite the need for further and wider studies, many recent reports suggest the presence of abnormalities in the cerebellum as well as other brain structures of ET patients.

Since postmortem studies provided evidence of degenerative changes in the cerebellum in many cases of ET, several imaging studies focused on visualizing cerebellar involvement in ET. Whereas some pioneer studies did not detect structural changes in the cerebellum, by magnetic resonance imaging (MRI) spectroscopy, a reduction in N-acetylaspartate/creatine and N-acetylaspartate/choline ratios was reported in cerebellar hemispheres (Louis et al. 2002; Pagan et al. 2003). Voxel-based morphometry studies using 1.5 T MRI (Daniels et al. 2006; Quattrone et al. 2008) led to controversial results; but recently, by means of 3 T MRI, structural white and gray abnormalities have been detected in ET patients in the cerebellum rather than cerebral areas. In particular, Benito-León et al. (2009) were able to report gray matter changes in the bilateral cerebellum and parietal and frontal lobes, and white matter changes in the frontal, parietal, and limbic lobes in 19 ET patients compared with 20 controls. Similarly, a recent report of 20 ET patients and 20 controls found scattered areas of cerebral and cerebellar atrophy (Bagepally et al. 2011); furthermore, a significant relationship between gray matter atrophy and tremor severity was observed in this study, suggesting that tremor generation is sustained by the degeneration rather than by dysfunction of cerebellar circuits. Further new insights came from diffusion tensor imaging studies: although a previous study did not find any significant difference between ET patients and controls (Martinelli et al. 2007), more recently, another report (Shin et al. 2008) showed a reduction of fractional anisotropy in the cerebellum and red nucleus. Again, by using 3 T MRI, significant cerebellar changes, as reduction of fractional anisotropy, were reported in patients with familial ET in the dentate nucleus and superior cerebellar peduncles (Nicoletti et al. 2010); moreover, a significant correlation between reduction of fractional anisotropy and ET disease duration was reported in this study. Finally, Klein et al. (2011), by means of a whole-brain analysis with tract-based spatial statistic, confirmed by voxel-wise analysis, have found evidence for localized pathology of cerebellar circuits and more general alterations of white matter in both motor and non-motor structures, supporting the impairment of neuronal fiber integrity in ET.

Progression of signs and symptoms in ET

The natural history of ET has not been completely defined, probably because of the heterogeneity of disease, but given the current evidence, three main observations can be made: age is a risk factor for ET; ET is a slowly progressive disorder; and clinical characteristics could differ at different stages of the disease. Moreover, to the best of our knowledge, there are no reported cases of spontaneous remission of ET.

As with either PD or AD, the incidence of ET is higher in elderly patients, ranging from 58.6 up to 84.3 per 100,000 individuals, which increases with age, suggesting that older age is an important risk factor for ET (Rajput et al. 1984). A few longitudinal studies have addressed the question of disease progression; an increase of tremor amplitude of about 30 % in a period of 4 years paralleled by a reduction in tremor frequencies has been detected by means of electrophysiologic evaluation (Elbe Elble 2000), and age and disease duration have been independently related to disease severity as assessed by a tremor scale (Louis et al. 2003). This evidence might suggest that a worsening of ET usually appears with disease duration, independent of age, and the higher rate of disease progression in older onset ET (Louis et al. 2009b) might indicate an age-related process of brain damage. A few longitudinal data (Louis et al. 2011b; Putzke et al. 2006) have defined a small, but significant, increase in tremor severity in ET patients estimated in terms of 12 % in a year from baseline evaluation in one series, and between 3.1 and 5.3 % in a recent paper assessing disease progression of tremor severity in ET patients followed up for many years. Moreover, it should be recalled that these changes in disease severity are more pronounced in elderly patients (Louis et al. 2009b). In addition, there are suggestions that despite a classic onset of tremor in both hands, later stages are characterized by the diffusion to other anatomic regions (mainly the head) (Louis 2005). Disease progression is not only characterized by the possible speading of tremor to the head or the increase in tremor amplitude, but also by the occurrence of signs of cerebellar involvement.

The neurologic disorders associated with ET include a wide spectrum of cerebellar symptoms, such as gait ataxia (Louis et al. 2010b, Rao et al. 2011), dysarthria (Kronenbuerger et al. 2009), oculomotor abnormalities (Helmchen et al. 2003), as well as deficits in hand–eye coordination (Trillenberg et al. 2006); however, these findings are not surprising since the possible presence of intention tremor among phenotypes of ET also represents a link between such disease and cerebellar dysfunction (Deuschl et al. 2000).

Furthermore gait ataxia, observed in 50 % of ET patients, is especially present in those individuals with longer disease duration and more severe tremor, and it is closely related to aging and tremor duration (Singer et al. 1994). The finding that cerebellar dysfunction is associated with ET in the advanced stages (Deuschl et al. 2000; Helmchen et al. 2003; Stolze et al. 2001) might indicate that the spreading of the disease to the cerebellum is a sign of progression, thus supporting the neurodegenerative hypothesis.

Klebe et al. (2005) have demonstrated that gait ataxia could be reversed in alcohol-responsive tremor patients by alcohol intake. Although this finding could be considered as evidence supporting the functional hypothesis, it should also be remembered that the neurochemical deficit sustaining ataxia and tremor could be the same, i.e. a deficit in gamma-amminobutyric acid (GABA) cerebellar neurotransmission; thus, tremor and ataxia could respond to GABAergic enhancement operated by ethanol, as in PD, most of the symptoms are responsive to dopamine replacement therapy.

A criticism has been made that these subclinical abnormalities are very mild and not clinically significant, but this observation does not invalidate their role in supporting a progressive cerebellar degeneration associated with ET.

Rest tremor has been described in 18.8 % of ET patients (Cohen et al. 2003); in a similar study, rest tremor was associated with more long-standing disease, more severe, and more disseminated tremor. Even if the substrate of resting tremor in ET is still unknown (Louis et al. 2011a), it could be due to pathologic spreading to extracerebellar circuits in patients with severe, long-standing, and disseminated disease.

Several papers have explored the cognitive profile of ET patients. Gasparini et al. (2001) have investigated the frontal lobe function in ET and PD patients, reporting that the ET group scored better on the tests than the PD group, but worse than the controls. Further studies have also shown that ET patients have impaired memory function (Lombardi et al. 2001) in addition to impaired executive function (Duane and Vermilion 2002). However, most of the neuropsychologic studies were limited due to the number of patients recruited, the lack of a control group, and the psychometric measures used were widely heterogeneous and not comparable. Despite these limitations, there is convincing evidence that ET patients have cognitive deficits, in particular, attention, executive function, verbal fluency, and several memory functions, including working memory, are mildly impaired in ET patients, whereas visuospatial function is relatively spared. The degree of cognitive impairment in ET is mild, related to age, and is sustained by a dysfunction in the dorsolateral prefrontal cortex and its connection to other brain areas, including the cerebral–cerebellar loop (Cerasa et al. 2010). Evidence has emerged, in a longitudinal study (Louis et al. 2010a), that the cognitive deficits in ET are not static, but seem to be progressing at a faster rate than occurs in normal elderly individuals, which supports the degenerative nature of ET.

The neuropsychologic evidence of a clear-cut impairment in many cognitive functions in ET patients is possibly due to impairment of the cerebello-thalamo-frontal connection as well as to direct impairment of other cortical and subcortical structures. This suggests that the clinical spectrum of symptoms in ET patients is to be widely expanded to other non-motor symptoms, as is the case for PD.