Introduction

Spinal and bulbar muscular atrophy (SBMA) is a rare, progressive motor neuron disease (MND) with multisystem involvement manifesting as androgen insensitivity, diabetes, sensory neuropathy, or autonomic nervous system involvement (Kennedy et al. 1968). The dominant phenotypic characteristics of the disease include muscle weakness and wasting of the limb muscles, bulbar muscles and respiratory muscles, muscle cramps, fasciculations, tremor, sensory neuropathy, breast enlargement, and erectile dysfunction (Kennedy et al. 1968; Fratta et al. 2014). The autonomic nervous system may be subclinically involved (Rocchi et al. 2011). SBMA is a microsatellite (trinucleotide) expansion disorder due to an enlargement of a CAG-repeat >38 in exon 1 of the androgen receptor (AR) gene on the long arm of chromosome X (Fratta et al. 2014; La Spada et al. 1991). Rarely, contracted repeats are associated with disease which is not SBMA (Kooy et al. 1999). Mouse models have shown that the CAG-repeat expansion is not sufficient for disease and that the presence of androgen is crucial to initiate disease development (Mariotti et al. 2000; Chevalier-Larsen et al. 2004; Schmidt et al. 2002; Katsuno et al. 2002).

SBMA is usually regarded as a disorder with adult onset (Mariotti et al. 2000) between the third and fifth decade of life (Vandenberghe et al. 2009). However, there is an increasing number of reports showing that in single cases SBMA starts already before adulthood (Sperfeld et al. 2002; Echaniz-Laguna et al. 2005; Ogata et al. 2001). Clinical manifestations (symptoms, signs, instrumental findings) at onset are highly variable between the studies, and it is under debate which of these is the most frequent and if age at onset is truly dependent on the CAG-repeat length if uncommon clinical or subclinical manifestations are included as initial manifestations. Considering classical onset motor manifestations, there is a clear negative relation between age at onset and the CAG-repeat size. The longer the CAG-repeat-expansion, the earlier the onset of SBMA (Fratta et al. 2014; Rhodes et al. 2009; Atsuta et al. 2006; La Spada et al. 1992; Xie et al. 2010) but CAG-repeat length is not associated with disease progression or the severity of the disease (Atsuta et al. 2006; Lu et al. 2009). However, <47 repeats is associated with a sensory dominant phenotype and leg tremor and >46 repeats with a motor-dominant phenotype (Nishiyama et al. 2014).

The present study aimed at answering the following questions: (1) Which clinical manifestations have been described so far in the literature as initial manifestations? (2) Which was the age at onset of these clinical manifestations? and (3) Is age at onset dependent on the CAG-repeat length if non-motor manifestations are additionally considered? Recognising the true onset of SBMA is crucial since potential treatment should start before neuronal degeneration evolves.

Methods

A systematic PubMed literature search was carried out by application of the search terms “bulbospinal muscular atrophy,” “spinal and bulbar muscular atrophy,” “Kennedy’s disease,” “motor neuron disease,” “weakness,” “tremor,” “cramps,” “fasciculations,” “twitches,” “breast enlargement,” “androgen receptor,” “CAG-repeats,” “poly-Q,” in combination with “human,” “onset,” “initial presentation,” “presenting manifestation,” and “symptom.” Additionally, references in these articles were searched for further papers suitable to be included in the discussion. Only studies in which the diagnosis was genetically confirmed were included.

Results

Upon application of the applied search terms, 221 hits were achieved (Fig. 1). Among these only 67 articles were assessed as suitable to answer the questions addressed in the introduction (Fig. 1). Twenty-two of the papers were only accessible as an abstract (Vandenberghe et al. 2009) and six were written in an inappropriate language (Fan et al. 2014; Yang et al. 2010). Sixteen of the included papers were cohort studies, 23 reported case series, and 12 were single case reports. The rest were reviews, editorials, or letters. The publication years ranged from 1968 to 2015.

Fig. 1
figure 1

Selection tree of the included papers

Full size image

The five largest clinical trials covering onset manifestations of SBMA are the ones by Atsuta et al. 2006, by Rhodes et al. 2009, by Fratta et al. 2014, by Mariotti et al. 2000, and by Sperfeld et al. 2002 (Fratta et al. 2014; Mariotti et al. 2000; Sperfeld et al. 2002; Rhodes et al. 2009; Atsuta et al. 2006). In the Japanese study of 223 patients, median age at onset of hand tremor (n = 126) was 33 years, of muscle weakness 44 years (n = 217), of dysarthria 50 years (n = 98), and of dysphagia 54 years (n = 79) (Table 1) (Atsuta et al. 2006). Non-motor manifestations were not assessed. Individual values of age at onset were provided only in a figure according to which tremor started between 20 and 52 years of age, weakness between 32 and 58 years, dysarthria between 35 and 64 years, and dysphagia between 36 and 65 years (Atsuta et al. 2006). All initial manifestations showed a wide range of individual onset age variability between 25 and 30 years (Atsuta et al. 2006). The number of CAG-repeats ranged from 40 to 57. Age at onset of evaluated manifestations correlated negatively with the CAG-repeat size (Atsuta et al. 2006).

Table 1 Studies of SBMA patients reporting type or frequency of onset manifestations
Full size table

In a study of 57 American SBMA patients, the most common presenting symptoms were muscle cramps (n = 22 (32 %)), tremor (n = 16 (23 %)), and leg weakness (n = 16 (23 %)) (Rhodes et al. 2009). More rarely, breast enlargement (7 %), upper limb weakness (7 %), and other non-specified manifestations were described (Rhodes et al. 2009). In half of the patients weakness occurred first in the lower limbs followed by bulbar weakness (33 %) (Rhodes et al. 2009). Age at onset was not provided for any of these manifestations but onset of muscle weakness ranged from 18 to 64 years (Rhodes et al. 2009). The number of CAG-repeats ranged from 41 to 53 (Rhodes et al. 2009).

In a study of 46 SBMA patients from the Great Britain, the most common initial manifestations were lower limb weakness (39 %) followed by upper limb weakness (10 %), tremor (6 %), bulbar weakness (3 %), breast enlargement (3 %), and muscle cramps (3 %) (Fratta et al. 2014). Age at onset of all onset symptoms ranged from 14 to 75 years (mean: 43.4 years) (Fratta et al. 2014). Age at onset of tremor ranged from 18 to 62 years (Fratta et al. 2014). Muscle weakness appeared between age 25 and 61 years. Erectile dysfunction appeared between age 38 and 63 years. Dysarthria started between 40 and 65 years of age (Fratta et al. 2014). Dysphagia had its onset between age 38 and 64 years. Breast enlargement occurred at a mean age of 39.1 years (Fratta et al. 2014). The first deficit noticed by patients in this study was hand tremor at a mean age of 35 years, followed by muscle weakness at a mean age of 43 years, dysarthria, and dysphagia, first recognized between 50 and 60 years of age (Fratta et al. 2014). Muscle cramps were noticed at a mean age of 39.5 years (Fratta et al. 2014). Muscle twitches were first reported at a mean age of 41.8 years (Fratta et al. 2014). Numbness or tingling was first reported at a mean age of 50 years (Fratta et al. 2014). CAG-repeat length ranged from 40 to 53.

In a study of 36 Italian patients, the authors provided individual values for age at onset and CAG-repeat size of all patients. Unfortunately, they did not consider non-motor onset manifestations. The 36 patients belonged to 30 unrelated families (Mariotti et al. 2000). Initial manifestations were lower limb weakness (n = 23), cramps (n = 6), upper limb weakness (n = 4), dysphagia (n = 1), and tremor (n = 1) (Mariotti et al. 2000). In four patients, the initial manifestations were not reported (Mariotti et al. 2000). Age at onset of these manifestations ranged from 25 to 60 years (Mariotti et al. 2000). Age at onset was not available in four patients. The number of CAG-repeats ranged from 39 to 50 (Mariotti et al. 2000).

In a study of 34 German patients, the initial manifestations were gynecomastia (52 %), premature exhaustion (28 %), tremor (26 %), cramps (16 %), myalgia (12 %), muscle weakness (4 %), or bulbar involvement (4 %) (Sperfeld et al. 2002). Unfortunately, age at onset of these abnormalities was provided only in a figure (Sperfeld et al. 2002). Number of CAG-repeats ranged from 41 to 52 (Sperfeld et al. 2002). The authors concluded that onset of SBMA is already in adolescence, thus earlier than commonly assumed (Sperfeld et al. 2002). In a study of 155 Chinese patients, onset was defined as onset of muscle weakness (Ni et al. 2015). Age at onset of muscle weakness ranged from 24 to 71 years (mean 44.2 ± 10.2 years) (Ni et al. 2015). Type and onset of non-motor manifestations was not provided. The CAG-repeat length ranged from 42 to 61 (Ni et al. 2015). In a study of 133 patients of whom 99 were in the placebo arm of a treatment study and 34 untreated control SBMA patients, age at onset ranged from 23 to 66 years (Hashizume et al. 2012). No further information about onset manifestations or CAG-repeat length was provided. In a study of 47 Japanese patients, examining tongue pressure age at onset ranged from 30 to 59 years (mean 44.4 ± 7.6 years) (Mano et al. 2014). Individual onset manifestations were not mentioned. The number of CAG-repeats ranged from 41 to 57 (mean 48.3 ± 3.3) (Mano et al. 2014). No individual values of these parameters were provided and no correlation between these parameters had been calculated (Mano et al. 2014).

A number of other studies, which included <25 patients or single patients and described onset manifestations, have been published (Table 1). In a study of 21 Taiwanese patients, mean age at onset of muscle weakness was 39 ± 7 years (Fu et al. 2013). Six patients noted hand tremor and two patients muscle cramps prior to onset of muscle weakness (Fu et al. 2013). Tremor and muscle cramps occurred between 15 and 36 years of age (mean 24 ± 8 years) (Fu et al. 2013). All patients developed bulbar manifestations at age 26 to 51 years (mean 39.7 years) (Fu et al. 2013). Among the 20 patients with muscle weakness, weakness began proximally in 70 % (Fu et al. 2013). CAG-repeat length ranged from 42 to 53 (Fu et al. 2013).

In a study of 18 Chinese patients from 16 families, age at onset ranged from 19 to 48 years (mean 31 years) (Hui et al. 2004). Initial manifestations were fatigability (39 %), tremor (22 %), focal weakness (17 %), cramps (11 %), and muscle twitching (11 %) (Hui et al. 2004). Non-motor manifestations were considered. CAG-repeat length ranged from 43 to 55 (Hui et al. 2004). In a study of 12 Chinese patients, limb weakness was the initial manifestation in all patients (Xie et al. 2010). There was no mentioning at which age muscle weakness occurred but mean disease duration at diagnosis was 14‚ years (Xie et al. 2010). The number of CAG-repeats ranged from 43 to 57 (Xie et al. 2010).

In a study of 11 French SBMA patients, 7 had unusual initial manifestations (Vandenberghe et al. 2009). These included myasthenia, cramps, fasciculation syndrome, polyneuropathy, post-traumatic monomelic neuronopathy, effort-dependent muscle intolerance, and/or muscular dystrophy (Vandenberghe et al. 2009). In a study of 11 Italian patients from the region of Reggio Emilia, mean age at onset was 44.8 ± 10.1 years (Guidetti et al. 2001). In a tremor study of 10 Brazilian SBMA patients from 7 families, age at onset was provided for each of the patients but, unfortunately, onset symptoms were not mentioned (Dias et al. 2011). Age at onset ranged from 19 to 49 years (Dias et al. 2011). The number of CAG-repeats was provided for each patient ranging from 41 to 53 (Dias et al. 2011).

In 10 Taiwanese SBMA patients participating in a FDG-PET study showing hypometabolism of the frontal lobes, age at onset ranged from 14 to 37 years (Lai et al. 2013). Mean age at onset was 27.3 ± 6.3 years. The number of CAG-repeats in these patients ranged from 46 to 56 (mean 49.9 ± 3.1) (Lai et al. 2013). In a study of 5 Peruvean patients, age at onset ranged from 31 to 63 years (Gomez-Calero et al. 2013). Manifestations at onset in these patients were tremor and cramps (n = 1), weakness and cramps (n = 3), or weakness exclusively (n = 1) (Gomez-Calero et al. 2013). The number of CAG-repeats varied between 43 and 49 (Gomez-Calero et al. 2013). In a study of 5 Italian SBMA patients, age at onset was 50, 43, 45, 36, and 28 years of age, respectively (Rocchi et al. 2011). The number of CAG-repeats was given as 43, 46, 43, 49, and 48, respectively (Rocchi et al. 2011). Unfortunately, the type of onset manifestations was not provided.

In a study of 3 Chinese patients from two families, age at onset of first clinical manifestations was 39, 39, and 41 years, respectively (Yang et al. 2010). The number of CAG-repeats amounted to 49, 48, and 47, respectively (Yang et al. 2010). Manifestations at onset were not provided. In a study of 3 German patients, the initial manifestations were symmetric proximal weakness, starting at age 39 years (patient A); muscle cramps in the lower limbs and symmetric proximal weakness in the upper as well as lower limbs starting at age 32 years (patient B); and symmetric proximal weakness of the upper limbs at age 45 years (patient C) (Kuhlenbäumer et al. 1998). The number of CAG-repeats was 51 in each of the three patients (Kuhlenbämuer et al. 1998). In a study of 2 Japanese brothers with SBMA, age at onset of muscle weakness was 66 and 78 years, respectively (Hemmi et al. 2009). The number of CAG-repeats was 42 in both of them (Hemmi et al. 2009). In the younger brother, gynecomastia appeared already in his 20s. Ostensibly, there was a clear negative correlation between onset and CAG-repeat size but considering gynecomastia as initial symptom, the correlation became non-significant (Hemmi et al. 2009). In an 11-year-old boy with SBMA and a CAG-expansion of 44, the patient presented with under-masculinized genitalia, microphallus, and scrotal hypospadia since age 4 years (Ogata et al. 2001).

Taken together, 16 different clinical manifestations were reported as onset manifestations of SBMA (Table 2). Age at onset of these manifestations ranged from 4 (Ogata et al. 2001) to 78 years (Hemmi et al. 2009). There was no consensus on the clinical presentation of SBMA at onset. Onset manifestations varied according to the definition of onset symptoms. Most studies regarded only motor manifestations as initial manifestation of the disease. Only in a few studies that non-motor manifestations were accepted as onset manifestations (Ogata et al. 2001). When correlating age at onset with CAG-repeat length by including data from studies which provided individual values and also included non-motor onset manifestations (Rocchi et al. 2011; Ogata et al. 2001; Yang et al. 2010; Hui et al. 2004; Lai et al. 2013; Gomez-Calero et al. 2013; Kuhlenbäumer et al. 1998; Hemmi et al. 2009), there was no significant correlation between age at onset and CAG-repeat length (r = 0.56) (Fig. 2). Differential diagnoses which had been considered before diagnosing SBMA included spinal muscular atrophy, amyotrophic lateral sclerosis (ALS), Parkinson’s disease, adrenoleucodystrophy, Becker muscular dystrophy, polyneuropathy, limb girdle muscular dystrophy, myasthenia, and hexosaminidase deficiency (Fratta et al. 2014; Vandenberghe et al. 2009; Kuhlenbäumer et al. 1998).

Table 2 Initial manifestations in SBMA reported in the literature
Full size table
Fig. 2
figure 2

Correlation (Pearson correlation coefficient) between age at onset and CAG-repeat length calculated from data taken from eight different studies, which provided individual figures and of which some also considered non-motor initial manifestations (Rocchi et al. 2011; Ogata et al. 2001; Yang et al. 2010; Fu et al. 2013; Dias et al. 2011; Lai et al. 2013; Gomez-Calero et al. 2013; Kuhlenbäumer et al. 1998)

Full size image

Discussion

This study shows that not only classical or motor manifestations should be considered as initial manifestations in SBMA but also more rare and less well-appreciated phenotypic features, as shown in Table 2. Particularly, subclinical and non-motor manifestations should be added to the list of onset manifestations. Secondly, this study shows that individual age at onset, type of manifestations at onset, and the frequency of onset manifestations are frequently not provided. Only ranges or mean or median values are available. Some studies provided age at onset but did not provide the typical onset manifestations or their frequency and vice versa. The previously propagated notion that age at onset is dependent on the CAG-repeat length has to be revised.

Initial manifestations of SBMA and age at onset may be missed because the history is not taken thoroughly enough, because certain manifestations, particularly rare or subclinical manifestations, are not regarded as onset manifestations of the disease, or because only motor symptoms are considered as onset manifestations of the disease (Mariotti et al. 2000; Tanaka et al. 2014). A further reason why onset of the disease may be missed could be a patient’s poor awareness of certain symptoms, such as tremor or easy fatigability, about which only unreliable information is provided. Exclusion of subclinical or non-motor manifestations is misleading, may delay diagnosis and symptomatic treatment, and may prevent early genetic counseling. There is currently no study available which assesses if creatine-kinase is elevated long before any motor manifestations develop or if androgen levels are decreased and estrogen levels increased long before any clinical endocrine manifestations occur. Subclinical parameters which could be assessed before appropriate clinical manifestations develop include tongue pressure, which has been shown to reflect swallowing function in patients with SBMA and appears to decrease before awareness of subjective dysphagia (Mano et al. 2014). Tongue pressure measurement could become a novel biomarker of SBMA, which is applicable for early-stage detection of the disease (Mano et al. 2014). Whether the increased size of the pituitary gland should be regarded as subclinical initial manifestation remains speculative (Pieper et al. 2013). Though 10 % of SBMA patients develop diabetes during the disease course (Kuhlenbäumer et al. 1998), it is unclear if diabetes can be an initial manifestation of the disease. Sensory symptoms because of polyneuropathy have been only rarely assigned as onset manifestations (Vandenberghe et al. 2009). Concerning the frequency of onset manifestations, there is a wide range of variability. Depending on which manifestations are regarded as onset manifestations, and if only motor manifestations or also non-motor manifestations are considered, the most frequent onset manifestations include weakness, cramps, tremor, gynecomastia, and fatigability (Table 1) (Fratta et al. 2014; Mariotti et al. 2000; Sperfeld et al. 2002; Rhodes et al. 2009; Hui et al. 2004).

The issue of the relation between the number of CAG-repeats and age at onset of clinical manifestations is still unsolved (Hemmi et al. 2009). This is due to inconsistency concerning the definition of what should or should not be considered as an onset manifestation. Most studies define onset of SBMA as muscle weakness first recognized by the patient (Mariotti et al. 2000; Tanaka et al. 2014). However, motor functions are frequently not the initial clinical manifestation why the correlation between age at onset and CAG-repeat length needs to be newly assessed by including more rare manifestations as has been tried in this study (Fig. 2). With regard to the correlation between motor functions and CAG-repeat length, there is no consensus in the literature. Various authors found a significant negative correlation between initial motor manifestations and CAG-repeat length (Mariotti et al. 2000; Atsuta et al. 2006; Igarashi et al. 1992; Shimada et al. 1995), whereas others did not (Mano et al. 2014; Guidetti et al. 1996; Amato et al. 1993; Biancalana et al. 1992; MacLean et al. 1995; Morrison et al. 1998). The significant negative correlation in some studies suggests that the pace of initial cellular events leading to motor neuron damage may be timed by the size of the genetic expansion in a CAG-repeat length-dependent manner, whereas the CAG-repeat length has only a limited predictive value for the severity of clinical impairment (Mariotti et al. 2000). Particularly, when all different types of onset manifestations were considered, the correlation between age at onset and CAG-repeat length became weak and non-significant (Fu et al. 2013). Hemmi et al. found that CAG-repeat expansions were not related to age at onset of androgen insensitivity but only to age at onset of motor functions (Hemmi et al. 2009). Unfortunately, there is no large study available, which includes and evaluates all different onset manifestations given in Table 2, including non-motor manifestations. The lack of a large study of this kind is responsible for the uncertainty about the correlation between CAG-repeat length and age at onset. However, calculating the correlation between age at onset and CAG-repeat expansion upon integration of all data available from studies providing individual data of any possible onset manifestation, it was not significant in the present study.

In conclusion, this study shows that age at onset depends on the definition of what shall be regarded as an onset manifestation. Non-motor and subclinical manifestations need to be recognized as onset manifestations as well. If unusual, non-motor manifestations or subclinical manifestations are additionally considered, age at onset is independent of the CAG-repeat length. SBMA patients need to be thoroughly investigated for their onset manifestations to treat these patients as early as possible in case effective measures become available in the future. Since life expectancy is hardly reduced in SBMA, it could be helpful to reinvestigate patients from previous studies with regard to their initial disease profiles.