Prevalence of fragile X‐associated tremor/ataxia syndrome: A survey of essential tremor patients with cerebellar signs or extrapyramidal signs

Abstract Objectives In screening studies of Western patients with cerebellar dysfunction, FMR1 premutations have been detected. A screening study of East Asian patients with presumed essential tremor (ET) did not detect these mutations, possibly because the ET patients did not closely mimic the phenotype of fragile X‐associated tremor/ataxia syndrome (FXTAS). The aim of this study was to estimate the prevalence of FMR1 premutations in a carefully recruited group of ET patients with additional phenotypic features of FXTAS. Materials and Methods From April 2014 to April 2018, we prospectively recruited patients with ET diagnoses from three tertiary care centers. Demographic and clinical data were collected, as well as data on presence of cerebellar signs and extrapyramidal signs (EPS). Tremor, cerebellar signs, and EPS were evaluated using appropriate clinical rating scales. For ET patients with additional cerebellar signs or EPS, FMR1 mutation analysis and brain magnetic resonance imaging were performed. Results Six hundred and three ET patients were recruited. Cerebellar signs or EPS were present in 168 (27.9%) of 603. FMR1 CGG repeat analysis was performed in 74 of 168 patients. Fifty‐two of 74 had cerebellar signs only, three had EPS only, and 19 had both neurologic abnormalities. Two patients had a FMR1 premutation and fulfiled both clinical and radiological criteria of FXTAS. Conclusions Two (2.7%) of 74 patients with presumed ET and additional neurological features were discovered to have FXTAS. The possibility of FXTAS should be considered in patients with ET who exhibit mild cerebellar signs or EPS.


| INTRODUC TI ON
Fragile X-associated tremor/ataxia syndrome (FXTAS) is a genetic disorder caused by an expanded CGG trinucleotide repeat located in the 5′ untranslated region of the fragile X mental retardation 1 (FMR1) gene. Fragile X syndrome results from a repeat number larger than 200 (i.e., the full mutation), while FXTAS results from a repeat number between 55 and 200, which is referred to as the premutation (Hagerman & Hagerman, 2013). The first cases of FXTAS were reported in 2001 (Hagerman et al., 2001). FXTAS primarily affects males, with a typical age of onset between 60 and 65 years (Hagerman & Hagerman, 2013). The main clinical features are slowly progressive intention tremor and gait ataxia. Other clinical features include peripheral sensory neuropathy, autonomic dysfunction, memory and executive function deficits, and extrapyramidal signs (EPS) (Hagerman & Hagerman, 2013). Various groups of movement disorder patients have been screened for FXTAS. As such, FMR1 mutations have been detected in groups of Western patients whose main clinical feature was progressive cerebellar ataxia (Brussino et al., 2005;Macpherson, Waghorn, Hammans, & Jacobs, 2003;Van Esch et al., 2005). Similar studies performed in Western (Clark, Ye, Liu, & Louis, 2015;Deng, Le, & Jankovic, 2004;Garcia Arocena et al., 2004) or Eastern (Tan et al., 2004) essential tremor (ET) patients did not detect FXTAS. It is worthy of note, however, that the ET patients in those studies did not closely mimic the phenotype of FXTAS.
Essential tremor is a disorder with a wide spectrum of clinical features. Classically, ET has been considered a disorder with a benign long-term course characterized by kinetic tremor without other neurological abnormalities. However, a growing number of studies now clearly demonstrate the coexistence of other neurological features. Patients with ET commonly exhibit cerebellar signs such as intention tremor or gait disorder (Arkadir & Louis, 2013;Deuschl, Wenzelburger, Loffler, Raethjen, & Stolze, 2000). Cognitive dysfunction (Lombardi, Woolston, Roberts, & Gross, 2001) or EPS (Thenganatt & Jankovic, 2016) have also been reported in ET patients. Of note, the postmortem examination of a woman with severe ET without parkinsonian features revealed Lewy bodies localized to the locus coeruleus, suggesting a connection between ET and Parkinson's disease (Louis et al., 2005). Currently, ET is considered a clinically heterogeneous disorder with numerous features of neurodegenerative disorders.
In fact, when ET is accompanied by such complex features (i.e., cerebellar signs, EPS), it can begin to resemble FXTAS. Indeed, patients with FXTAS have been assigned erroneous initial diagnoses of ET, before genetic analysis detected an FMR1 premutation (Hall et al., 2005). These examples suggest that FXTAS is sometimes so similar to ET that it is difficult to distinguish the two entities on a clinical basis. To determine the prevalence of FMR1 mutations in ET patients with cerebellar signs or EPS, we performed FMR1 mutation analysis in ET patients who shared these clinical features of FXTAS.  (Deuschl, Bain, & Brin, 1998), except for one of the exclusion criteria requiring absence of other abnormal neurologic signs.

| Subject selection
All enrolled patients had postural or kinetic tremor involving hands and forearms. Other neurologic abnormalities could be present, but they were only mild in all enrolled subjects and postural or kinetic tremor was the most prominent feature.

| Clinical assessment and workup
Demographic data, age of tremor onset, presence of first-degree relatives affected by tremor, and tremor response to alcohol were collected by history. Presence of rest, postural, or kinetic tremor, anatomic locations affected by tremor, and presence of cerebellar signs or EPS were determined by neurologic examination.
We excluded subjects who were confirmed with gene test to have autosomal dominant or recessive cerebellar ataxia syndromes other than FXTAS. We also excluded subjects who had bradykinesia and one or both of rest tremor and rigidity, which is the essential criterion of Parkinson's disease defined by the 2015 Movement Disorder Society criteria (Postuma et al., 2015). If EPS was progressive over time, those subjects were also excluded. In addition, we excluded subjects with prominent autonomic symptoms or signs and cerebellar sign or EPS suggestive of multiple system atrophy. We also excluded subjects with drug-induced tremor, structural brain lesion or history of brain surgery, and severe medical problem that could produce tremor or gait disturbances, including endocrine disorders, hepatic disease, uremia, and severe joint problems.
Tremor was assessed by the revised version of the clinical rating scale for tremor (CRST) developed by Fahn and coworkers in 1993. Scale for the assessment and rating of ataxia (SARA) and Unified Parkinson's Disease Rating Scale (UPDRS) part 3 (Fahn & Elton, 1987) scores were also assigned to each patient. These scores were assigned by movement disorder neurologists (Jin Whan Cho and Jinyoung Youn).
From the 603 recruited ET patients (Figure 1), we selected 168 patients who had cerebellar signs or EPS. Cerebellar sign was considered to be present when one or more of the following features was present: dysmetria or intention tremor during the finger-to-nose test, dysmetria during the heel-to-shin test, dysdiadochokinesia (hand) during the rapid alternating movements test, impaired tandem gait, and hypermetric saccades. EPS was considered to be present when at least one of the following signs was present: bradykinesia with decremental response, lead-pipe rigidity, or tremor at rest. If the patient agreed, we performed FMR1 CGG repeat number analysis and obtained a brain magnetic resonance image (MRI). The MRI was reviewed qualitatively for presence of middle cerebellar peduncle (MCP) or corpus callosum splenium (CCS) signs, both known to be useful diagnostic markers for FXTAS. All MRI scans were assessed by two experienced movement disorder neurologists, and a radiologic sign was thought to be present only when both agreed upon it.

| Genetic analysis
The FMR1 premutation screening was performed using a commercialized CGG repeat primed PCR (Asuragen Inc., Austin, TX). Testing was conducted based on the manufacturer's recommendations, as previously described (Chen et al., 2010). In this assay, full-length and CGG repeat primed amplicons were produced using two flanking gene-specific primers and one CGG repeat primer and were analyzed by capillary electrophoresis.

| Statistical analysis
We compared ET patients with versus without additional features (cerebellar signs or EPS) using chi-squared tests, Fisher's exact tests, or Student's t tests where appropriate (Table 1). For the presence of an affected first-degree relative or alcohol response, possible answers were "yes", "no", or "unknown" and statistical analysis was performed based on a 2 × 3 contingency table.
In addition, we divided the subjects who underwent FMR1 analysis into three groups: those with cerebellar signs and no EPS ("cerebellar signs group"), with EPS and no cerebellar signs ("EPS group"), and with both of cerebellar signs and EPS ("mixed group").
Demographic data, clinical data, CRST score, and presence of abnormal MRI signs were compared between all three groups using analysis or variance or Kruskal-Wallis H test according to whether the data followed a normal distribution. SARA score was compared between EPS group and one of cerebellar signs group or mixed group. UPDRS part 3 score was compared between cerebellar dysfunction group and one of EPS group or mixed group.

| D ISCUSS I ON
In contrast, we recruited only ET patients with additional mild cerebellar signs or EPS, which are important clinical features of FXTAS.
We found two cases of FXTAS from 74 patients (2.7%), which is a high proportion considering the low prevalence of FMR1 premutation in Korea. Several population-based studies in Korea investigated the prevalence of FMR1 premutation carriers by utilizing data on women of reproductive age and estimated it to be between 1/788 and 1/1090 (Han et al., 2012;Jang et al., 2014). This is lower than previously reported prevalence of FMR1 premutation carriers in Western countries. In the female population, it was 1/259 in a Canadian study (Rousseau, Rouillard, Morel, Khandjian, & Morgan, 1995). In the male population, the estimated prevalence ranged between 1/400 and 1/813 in Caucasians (Dombrowski et al., 2002;Hantash et al., 2011).
The prevalence of FXTAS in the screened patients of our study was 2.7% (2/74). This is comparable to the pooled prevalence of FXTAS in male cerebellar ataxia patients, which is 1.5% (16/1049), reported in a meta-analysis of FMR1 premutation screening studies (Jacquemont, Leehey, Hagerman, Beckett, & Hagerman, 2006 screening studies in ET did not detect any cases of FXTAS. Our study provided a positive result of screening for FMR1 premutation in patients whose main clinical picture was ET with additional mild cerebellar signs or EPS. It could help selecting future candidates for testing of FMR1 premutation. Middle cerebellar peduncle sign is a highly specific radiologic marker for FXTAS, with a reported sensitivity of 0.52 and specificity of 0.87 (Jacquemont et al., 2003). CCS sign is another imaging marker, defined by hyperintensity in the splenium of the corpus callosum. One study reported its sensitivity and specificity in men to be F I G U R E 3 Magnetic resonance images of patients with FMR1 premutation. Panels (a) and (b) are images of patient 1, and panels (c) and (d) are images of patient 2. Panels (a) and (c) shows corpus callosum splenium sign (red arrows), and panels (b) and (d) shows middle cerebellar peduncle sign (red arrowheads) From our study subjects, the two most common FMR1 CGG repeat numbers were 30 and 29. Patients corresponding to those two numbers accounted for most of our study subjects (77.0%). Thirty and twenty-nine are also the most common repeat numbers in a larger sample from the Korean population and in populations from other countries (Jang et al., 2014;Peprah, 2012). Therefore, the genetic distribution with regard to FMR1 in our study subjects likely represents the general population.
Mean of UPDRS part 3 score was significantly higher in the EPS or mixed group compared to the cerebellar signs group, which reflected our patient grouping scheme. However, mean of SARA score did not differ significantly between cerebellar signs or mixed group and EPS group. Nevertheless, the mean value was the smallest in the EPS group, and it was marginally significant. This lack of statistical significance may be due to the very small number of subjects

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.