Introduction

Spinal and bulbar muscular atrophy (SBMA), also known as Kennedy’s disease, is a rare X-linked hereditary neuromuscular disorder (~ 1–2 per 100,000)1 associated with a CAG trinucleotide repeat expansion in the first exon of the androgen receptor (AR) gene2. Typically, the clinical disease course manifests during adult life, presenting wasting and weakness of bulbar and peripheral extremities3,4. Although patients can demonstrate significant disability in later years, including respiratory involvement, lifespan is typically unaffected4. The slowly progressive nature of the disease course of SBMA poses a significant challenge for therapeutic development and clinical monitoring due to a notable absence of objective, clinically feasible, markers sensitive to short-term (< 12 months) disease-related change.

The primary mechanism of SBMA pathology relates to a toxic gain of function in mutant androgen receptor (AR) protein3. Aggregated AR proteins induce dysfunction and subsequent death of motor neurons resulting in a significant loss of motor neurons throughout all spinal cord segments and selective brainstem motor nuclei (i.e., hypoglossal, trigeminal), and to a lesser extent sensory neurons5. The primary lower motor neuron pathology underlying SBMA has been extensively investigated from a clinical and neurophysiological perspective1,3,4,6,7,8. Significant motor and sensory nerve dysfunction are evident, involving both upper and lower limbs8. Similarly, quantitative skeletal muscle imaging of fat infiltration in bulbar and limb muscles indicate marked abnormalities correlated with severity of functional motor impairment9. Increasingly, however, multiorgan involvement beyond muscle dysfunction (i.e., neuronal, autonomic, metabolic, hormonal, urinary) have been highlighted as potential therapeutic targets for improving clinical management and monitoring disease progression in SBMA10. Notably, multimodal evidence indicates the presence of significant central brain changes impacting structure, metabolism, and functional activation (Table 1). It is pertinent to highlight that considerable variability in central brain changes is likely present across independent SBMA patient cohorts11,12, given their presumed role as a secondary pathological mechanism of disease, developing over the life-long disease course.

Table 1 Summary of studies investigating central brain abnormalities in SBMA patients.
Full size table

Previous studies of central and peripheral nerve dysfunction in SBMA have tended to be conducted independently, such that any inter-relationship remains unclear. Furthermore, existing brain imaging studies in SBMA have been exploratory in nature, limiting the precision for capturing fine-grained patterns of disease-related motor degeneration, the primary feature of disease. Pathophysiological understanding of central neurodegeneration, beyond the brainstem and spinal cord, in SBMA is significantly less understood and has yet to be comprehensively examined in the context of motor and sensory dysfunction.

To determine the potential involvement of central circuitry linked to neurodegeneration in SBMA, the current study comprehensively examined function and structure across central and peripheral motor systems, any potential inter-relationship, and association with motor impairment in a cohort of clinically well-defined SBMA patients. Focus was placed on characterizing selective patterns of disease-related structural brain changes associated with the extended motor cortices (premotor, primary motor, somatosensory) and their respective corticospinal white-matter pathways of signal transmission. Drawing upon extensive evidence of cortical motor reorganization as a consequence of long-term peripheral limb injury13 and widespread neural pathology14,15, we hypothesized that chronic maladaptive motor and sensory responses will translate into quantifiable brain changes in SBMA. This study aims to expand the current understanding of the multi-system clinical profile of long-term SBMA pathophysiology.

Methods

Participants

Twelve genetically confirmed patients (defined as CAG repeat > 38) diagnosed with SBMA were prospectively recruited from specialist neuromuscular referral clinics across Australia (Royal Prince Alfred Hospital; Prince of Wales Hospital; Concord Hospital; Royal Brisbane and Women’s Hospital) between 2020 and 2023. All SBMA patients attended the Brain and Mind Centre and underwent comprehensive functional and electrophysiological assessment of peripheral limbs, and same day whole-brain magnetic resonance imaging (MRI) scan. Clinical upper motor neuron features were not observed in SBMA patients. Demographically matched ALS (N = 8) and healthy control (N = 7) participant cohorts were independently recruited from the motor neuron disease imaging biomarker database at the Brain and Mind Centre, which employed the same MRI protocol. All ALS participants presented with apparently sporadic, classical forms of ALS without dementia. Healthy control participants demonstrated no prior history of mental illness, head injury, movement disorders, alcohol and drug abuse, or abnormality on MRI. The study was approved by the University of Sydney human research ethics committees (2021/283; 2021/ETH00559), all patients provided written informed consent prior to participation. Research was conducted in accordance with the Declaration of Helsinki ethical guidelines.

Clinical and neurophysiological assessments

An existing battery of clinical assessments to interrogate peripheral motor and sensory nerve dysfunction was administered to SBMA patients. Functional assessments included: (i) revised amyotrophic lateral sclerosis functional rating scale (ALSFRS-R)16—a four domain motor disability questionnaire [0–48], with lower scores reflecting greater disability, and disease progression rate calculated as the number of points lost per month of disease [(48-ALSFRS-R/disease duration); (ii) hand grip strength—assessed using a dynamometer (Saehan Corp, Korea) in kilograms, patients were asked to grip the dynamometer as hard as possible for 5 s for each hand, repeated 3 times and average reported across trials; (iii) grooved pegboard test17—timed assessment of motor dexterity involving the placement of 25 pegs into grooved holes using their dominant hand, repeated across 2 trials and average time taken (seconds) reported; (iv) fingertip tactile spatial acuity (dominant index finger)18—quantified using the grating orientation task (JVP Domes, Stoelting Co.), dome gratings between 0.75 mm and 12 mm were placed onto the index finger either proximal-distally or lateral-medially in random order to identify the smallest grating that could be reliably discriminated (20 attempts), with lower widths indicating greater discrimination sensitivity; (v) nerve conduction—motor (compound muscle action potential [CMAP]) and antidromic sensory (sensory nerve action potential [SNAP]) nerve conduction of the right median (upper limb motor: abductor pollicis brevis, stimulation at wrist; upper limb sensory: second digit, stimulation at wrist), sural (lower limb sensory: lateral malleolus, stimulation 10–15 cm proximal) and tibial (lower limb motor: abductor hallucis, stimulation at medial malleolus) nerves19; (vi) Total Neuropathy Score clinical version (TNSc, Johns Hopkins University)20—widely used six domain composite tool [0–24] to assess severity of motor and sensory neuropathy incorporating clinical signs and symptoms, with higher scores reflecting greater neuropathy. SBMA patient performance across clinical assessments were interpreted relative to published normative control ranges.

MRI acquisition

Whole-brain imaging was performed using a 3 T MRI scanner (GE MR750, DV29; 32-channel Nova head coil). Coronal T1-weighted images were acquired using an MPRAGE sequence (TE/TR = 2.3/6.2 ms; flip angle = 12°; matrix size = 256 × 256; 1 mm3 isotropic; 204 slices). Diffusion-weighted images (DWI) were acquired using a HyperBand blipped-CAIPI sequence (140 directions; b-values = 700/1000/2800 s/mm2, 8xb0 interleaved; TE/TR = 100/3245 ms; flip angle = 90°; matrix size = 128 × 128; 2 mm3 isotropic; 66 slices; multi-band factor = 3). Three additional b0 volumes were acquired with the phase encoding polarity reversed. The MRI scan from one SBMA patient was excluded from analysis due to significant motion.

Volumetric analysis

T1-weighted images were processed using the standard recon-all pipeline in FreeSurfer V6.0. Briefly, images underwent correction for motion and intensity inhomogeneity, removal of non-brain tissue, and segmentation of white and gray matter. All segmentations underwent visual inspection for motion artifacts and proper gray/white matter tissue boundaries. Intra-cranial volume was calculated for all participants and used to correct for inter-individual differences. Mean cortical volume and thickness values associated with Brodmann labels of the premotor (BA6), primary motor (BA4), and somatosensory (BA1/2/3) cortices were quantified for all participants. Each individual’s T1-weighted image was linearly co-registered to native DWI space. Output transformation matrices were saved and applied to co-register respective cortical masks of Brodmann defined motor and somatosensory cortices.

Tractography

DWI images were pre-processed using MRtrix3. Briefly, images underwent denoising, eddy current correction21, N4 bias field correction22, and global intensity normalization using the median white matter value of the b0 image. Next, 3-tissue response functions representing single-fiber white matter (WM), grey matter, and cerebrospinal fluid were estimated from the data. Each participant’s averaged tissue specific response functions were used to estimate fiber orientation distribution for each voxel with the multi-shell multi-tissue constrained spherical deconvolution (MSMT-CSD) algorithm23. Anatomically constrained whole-brain tractography was then performed using default parameters to compute 10 million probabilistic streamlines24. Each streamline was assigned a weight, computed using CSD informed filtering of tractograms (SIFT2)25.

Corticospinal WM tracts associated with the premotor (BA6), primary motor (BA4), and somatosensory (BA1/2/3) cortices were delineated using their respective co-registered Brodmann labels as seed regions. A region of interest mask at the level of the pons was manually delineated on native DWI images as an inclusion waypoint. Masks of the cerebellum and corpus callosum were included as exclusion waypoints. All delineated tracts were visually inspected for accuracy. The diffusion tensor metrics of each WM tract were quantified across participants. Only the juxtacortical tract region, between cortical WM to the inferior boundary of the lateral ventricles, were assessed due to significant intra-voxel overlap of the corticospinal WM tracts beyond this point.

Statistical analyses

Statistical analyses examining differences in demographic, clinical, and quantitative MRI metrics, were performed between participant cohorts using SPSS V.21. Data normality was determined using the Shapiro–Wilk test. Analysis of variance (ANOVA) was used to examine group differences across demographic and clinical variables. Analysis of covariance (ANCOVA) was performed to examine group differences across MRI metrics, controlling for age. Planned two-tailed comparisons of premotor, motor, and somatosensory associated cortical and WM changes were performed between participant cohorts. Results were adjusted for multiple comparisons using Bonferroni correction. Pearson’s correlations were used to examine clinical correlations between CAG repeat expansion, and central and peripheral measures of motor and sensory integrity. P-values < 0.05 were considered significant.

Ethics approval

Ethical approval for this study was obtained from The University of Sydney Human Research Ethics Committee (2021/283; 2021/ETH00559). Research was conducted in accordance with the Declaration of Helsinki ethical guidelines.

Results

Patient demographics and clinical profile

Participant cohorts were demographically matched for age, gender, and handedness (Table 2; p-values > 0.57). All SBMA patients were genetically screened, demonstrating an abnormal CAG repeat length ranging between 41 and 46 (mean = 43.8, SD = 1.9) in the AR gene. Mean disease duration for SBMA patients was 9 years from initial clinical diagnosis. Mean disease duration for the disease control ALS cohort was 14 months from initial patient reported symptom onset. SBMA and ALS cohorts were well-matched for general functional motor impairment on the ALSFRS-R (p = 0.82). Expectedly, the ALS cohort demonstrated significantly greater disease progression rate, as reflected by change in ALSFRS-R score over the course of disease, relative to the SBMA cohort (p = 0.04).

Table 2 Demographic characteristics and clinical profile of ALS and SBMA patients, and healthy control participants.
Full size table

SBMA patients demonstrated marked changes across motor and sensory assessments, relative to published normative values (Table 3)19,26,27,28, with prominent sensory dysfunction of upper limbs appearing consistently across patients. Overall bilateral hand grip strength was reduced in SBMA (Left: mean = 14.5 kg, SD = 5.2; Right: mean = 19.5 kg, SD = 10.2) and was significantly abnormal in half of patients (left: 58%; right: 50%)26. Overall motor flexibility of the dominant hand was similarly impaired in SBMA and significantly abnormal in 58% of patients, as indicated by delayed completion time on the grooved pegboard task (mean: 93.9 s, SD = 16.4)27. Sensory tactile spatial acuity of the dominant index fingertip was consistently found to be abnormal in 92% of SBMA patients (mean = 6.9 mm, SD = 2.4)28. Electrophysiological assessment of motor nerve conduction indicated overall reduced CMAP of upper and lower limbs, which was abnormal in 42% of SBMA patients (Upper: mean = 7.1 mV, SD = 3.1; Lower: mean = 4.3 mV, SD = 1.6)19. All SBMA patients demonstrated significantly abnormal sensory nerve conduction of the right upper limb, as indicated by significantly reduced SNAP (mean = 10.5 μV; SD = 8.1)19. Sensory nerve conduction of the right lower limb was reduced in SBMA (mean = 3.6 μV; SD = 2.2), but SNAP was only significantly abnormal in 17% of patients, a likely underestimation based on high normative range variability and currently employed classification criteria19.

Table 3 Functional motor and sensory assessments, and nerve conduction.
Full size table

Cortical grey matter integrity

Integrity of grey matter volume and cortical thickness of premotor, motor, and somatosensory cortices were independently assessed between participant cohorts, across left and right hemispheres (Supplementary Fig. S1). Groupwise differences in cortical grey matter integrity did not survive multiple comparisons correction.

Corticospinal tract integrity

Corticospinal tracts traversing premotor, primary motor, and somatosensory cortices to the brainstem were reliably delineated in the left and right hemisphere across all participants. Spatial profiles of the respective tracts remained relatively distinct between the juxtacortical region to the inferior boundary of the lateral ventricles and used to compare altered WM diffusivity (Fig. 1). Relative to controls, SBMA demonstrated widespread WM abnormality across all delineated tracts, consistent across left and right hemispheres (p-values < 0.05). Increased axial diffusivity was observed in the premotor (left, p = 0.01; right, p = 0.02), primary motor (left, p = 0.04; right, p = 0.05), and somatosensory (left, p < 0.01; right, p = 0.04) tracts. No significant group differences were observed in fractional anisotropy, mean diffusivity, or radial diffusivity. Selective abnormality in axial diffusivity suggests the presence of primary axonal degeneration as the cause of central WM pathology in SBMA.

Figure 1
figure 1

Diffusivity of corticospinal tracts were assessed between cortical grey/white matter boundary to the level of the inferior boundary of the lateral ventricles, at which point corticospinal tracts demonstrated marked intra-voxel overlap. Axonal degeneration was observed within juxtacortical white-matter of corticospinal tracts (premotor; motor; somatosensory) in ALS and SBMA patients, relative to healthy control participants. Representative 3D reconstruction of delineated tracts from a healthy control participant is shown in native space. SBMA patients demonstrated widespread increase in axial diffusivity across all reconstructed corticospinal tracts across left and right hemispheres. ALS patients demonstrated a selective increase in axial diffusivity specific to corticospinal streamlines seeded from the primary motor and somatosensory cortices. Mean and standard deviation. *Indicates significance at p < 0.05; **indicates significance at p < 0.01; Bonferroni corrected.

Full size image

In contrast, ALS patients demonstrated a selective abnormality associated with juxtacortical WM of the primary motor and somatosensory corticospinal tract, relative to controls (left, p = 0.05; right, p = 0.05). No significant WM differences were observed between SBMA and ALS patient cohorts.

Clinical correlations

The severity of central WM degeneration in SBMA demonstrated widespread correlations with TNSc in SBMA (Fig. 2). Significant positive correlations were consistently found in the left and right hemisphere between increased axial diffusivity of juxtacortical WM across premotor (Left: R2 = 0.53, p = 0.02; Right: R2 = 0.45, p = 0.03), primary motor (Left: R2 = 0.58, p = 0.01; Right: R2 = 0.46, p = 0.03), and somatosensory (Left: R2 = 0.67, p < 0.01; Right: R2 = 0.38, p = 0.05) tracts, and increased TNSc. Similarly, increased disease duration demonstrated a significant positive correlation with increased peripheral neuropathy as reflected by increased TNSc score in SBMA patients (R2 = 0.85, p < 0.01), but not with age (R2 = 0.24, p = 0.15). With regard to increased disease progression rate, significant negative correlations were consistently observed in the right hemisphere with reduced fractional anisotropy of juxtacortical WM across primary motor (R2 = 0.44, p < 0.01) and somatosensory (R2 = 0.28, p = 0.05) tracts, as well as reduced cortical grey matter volume of the primary motor cortex (R2 = 0.32, p = 0.04. These findings support the interpretation that central neurodegeneration in SBMA accompanies disease progression.

Figure 2
figure 2

Increased white matter abnormality along corticospinal tracts (premotor; motor; somatosensory) in SBMA patients was significantly correlated with total neuropathy score clinical version (TNSc).

Full size image

While consistent correlations were observed between disease duration, central WM degeneration, and TNSc score in SBMA, no significant correlations were observed between standalone functional motor and sensory assessments or upper/lower limb nerve conduction (ALSFRS-R, CMAP, SNAP). Similarly, examination of ALSFRS-R ‘bulbar’ and ‘limb’ specific sub-domains did not demonstrate a significant correlation with WM integrity of corticospinal tracts in SBMA. Furthermore, no significant correlations were found between length of CAG repeat expansions in the current SBMA cohort with either central or peripheral assessments.

Discussion

The current findings suggest the presence of secondary neurodegenerative pathology manifesting during the disease course of SBMA, potentially impacting the integrity of central corticospinal pathways. Axonal degeneration of juxtacortical white matter underlying premotor, motor, and somatosensory corticospinal tracts was consistently altered across brain hemispheres in SBMA and demonstrated an association with severity of peripheral neuropathy suggesting this is an adaptive change. This profile of corticospinal white matter abnormality resembled that of the demographically matched ALS cohort, which clinically manifests upper and lower motor neuron dysfunction. These findings hold important implications for (i) considering the multi-system impact of chronic SBMA pathophysiology beyond lower motor neuron dysfunction, and (ii) the timescale of underlying SBMA disease activity prior to clinical symptom onset.

Evidence of central brain dysfunction in SBMA, beyond lower motoneurons in the brainstem, have been reported across independent neuroimaging and electrophysiology modalities and patient cohorts11,29,30,31,32,33, albeit not without contrasting reports of intact cortico-motor neuronal function12,34. A factor that has limited the significance of central brain abnormalities in the disease course of SBMA is the lack of meaningful clinical interpretation. The bulk of SBMA neuroimaging studies in the literature adopted exploratory whole-brain analyses with roughly half reporting an absence of significant correlation with clinical variables11,12,29,31,33. Studies employing a targeted study design have revealed a more interesting pattern of meaningful brain changes. Notably, gross structural brain abnormalities in a rare case of genetically confirmed SBMA with concomitant frontotemporal dementia35, reduced functional representation of cortical sensorimotor response to peripheral tactile stimulation36, and potential mediating influences of increased CAG repeat length with pituitary hormonal regulation37. Taken together with current findings of a consistent correlation between juxtacortical motor/somatosensory white matter abnormality and disease duration, there is diverse evidence supporting the view that central neurodegeneration is a significant component of multi-system dysfunction accompanying the clinical lower motor neuron syndrome of SBMA.

A notable finding arising from the current study was the correlation between disease duration and severity of central (axial diffusivity) and peripheral (TNSc), primarily sensory, dysfunction. As mentioned earlier, clinical associations of disease with reported brain abnormalities in SBMA have been difficult to interpret in a concordant manner30,38,39. The exception being, CAG repeat length does not significantly impact central white matter integrity11,29,38. Previous clinical associations raised by Garaci and colleagues are consistent with current imaging findings. Specifically, they demonstrated a clinical correlation between severity of functional motor impairment on the ALSFRS-R and generalized whole-brain white matter integrity, as well as a selective association between axonal degeneration of juxtacortical corticospinal white matter and disease severity38. Consistent with the view of central neurodegeneration being a disease feature of SBMA, this relationship was observed across juxtacortical white matter across left and right brain hemispheres. The current findings extend the clinical understanding of SBMA by demonstrating a consistent relationship between degeneration of corticospinal white matter and disease duration, directly highlighting the potential long-term impact of motor pathology beyond lower motor neuron dysfunction.

Both the underlying mechanisms and extent of central neurodegeneration, beyond the brainstem, are poorly characterized in SBMA. Harmonizing currently reported central white matter abnormalities in SBMA with known clinicopathological features is challenging. Firstly, clinical signs indicative of upper motor neuron lesions was not observed in SBMA3,4. Clinical correlations with diffusion metrics of premotor, motor, and somatosensory corticospinal tract white matter did not demonstrate an association with specific domains of motor dysfunction, only the composite measure of overall peripheral neuropathy (i.e., TNSc score) and ALSFRS-R disease progression rate. With regard to disease mechanism, the widespread and largely symmetrical hemispheric cortical white matter abnormality presents a stark difference to the highly selective involvement of brainstem motoneuron nuclei5. This likely reflects important divergent mediating factors of the impact of mutant AR on different neuronal populations. When taking into consideration the widespread distribution of mutant AR nuclear inclusions throughout deep-grey subcortical brain structures (i.e., thalamus, caudate, striatum) in SBMA15, and their extensive cortical circuitry40, it is plausible for white matter abnormalities resulting from long-term impaired neuronal development processes to emerge. This is consistent with recent multi-centre evaluations of blood biomarkers in SBMA demonstrating normal levels of peripheral neurofilament light and heavy levels, suggesting an absence of ‘active’ neuronal damage41,42. Currently reported patterns of corticospinal white matter abnormality represent a secondary rather than primary pathological feature of disease, highlighting central brain abnormality as an insidious indicator of SBMA pathogenesis.

An interesting direction for prospective investigation is reconciling somewhat conflicting findings from historical transcranial magnetic stimulation (TMS) in SBMA and neuroimaging findings. Previous findings by our group reported an absence of cortical motor dysfunction in an independent cohort of SBMA patients across threshold-tracking TMS parameters, including motor evoked potential, cortical silent period, and short-interval intracortical inhibition, concluding that cortico-motor neuronal circuits were intact in SBMA, validating TMS as a disease specific diagnostic marker of ALS34. While the clinical distinction with ALS is not disputed, current clinical correlations between central white matter abnormalities and disease duration suggests longitudinal follow-up is pertinent to examine whether the TMS signature of disease manifests in the later stages of SBMA disease course34. Particularly given the high sensitivity TMS holds in detecting sub-clinical dysfunction of cortico-moto neuronal circuits43. Structurally, the motor cortex in SBMA appears relatively ‘normal’ even at post-mortem14,15. Evidence for cortical motor dysfunction in SBMA has been somewhat demonstrated by magnetic resonance spectroscopy findings of N-acetyl-aspartate and myo-inositol metabolite abnormalities in patients30,33. Given the widespread axonal corticospinal degeneration observed in current analyses, physiological abnormalities on TMS would be expected unless there is a preferential degeneration of ‘somatosensory’ corticospinal fibers. There is, unfortunately, no method for discerning this through conventional MRI. Such a dissociation would, however, provide the neural basis for early neurophysiology studies in SBMA examining sensory multi-modal evoked potentials, reporting evidence of significant physiological abnormalities, namely prolonged peak latencies in somatosensory evoked potentials32. Contrasting modality findings could also simply represent differences in ‘localized’ brain changes captured on MRI and more ‘global’ physiological changes being captured by TMS. Reconciling the temporal association between neurophysiological responses and anatomical brain integrity in SBMA could potentially advance current understanding of central and peripheral mechanisms driving evolving peripheral dysfunction44.

There is growing recognition of the utility of quantitative MRI as sensitive objective biomarkers for capturing disease trajectory in SBMA3. Novel biomarkers sensitive to the impact of disease related activity over practical drug evaluation periods are critical for therapeutic development. This is of particular concern given the insensitivity of primary outcomes measures used across existing SBMA clinical trials, and the pipeline of at least 30 therapeutic approaches demonstrating efficacy in transgenic mice currently progressing towards human translation3. While a longstanding focus has been placed on skeletal muscle41, current findings suggest longitudinal validation of central brain changes in relation to peripheral dysfunction is perhaps warranted to advance current understanding of the holistic impact of SBMA pathology and explore new avenues for quantitative MRI biomarker development. Comparative longitudinal studies of male and female genetic carriers are of significant interest, potentially identifying testosterone-independent neurodegenerative mechanisms in female SBMA carriers presenting with mild neuromuscular symptoms45. As with all single-centre studies in SBMA, patient sample size was limited, with patients presenting with variable lengths of disease duration, and clinical assessments performed at a single time-point. Clinical motor/sensory assessments were also directly translated from established research protocols employed at our clinic. A collaborative multi-centre approach will be essential for validating imaging metrics given the significant disease heterogeneity arising from rare neurodegenerative disorders, including consensus on harmonized clinical assessments for SBMA. A larger SBMA patient sample size would allow further interrogation of concomitant medical issues and their potential contribution to structural brain abnormalities, such as high blood pressure and diabetes. The scope of the current study was limited to examining the association between central MRI and established clinical assessments of peripheral dysfunction, targeting cortical motor and sensory pathways. Fine-grained tractography based delineation of selective brainstem motor nuclei dysfunction remains noticeably absent from the SBMA literature and offers a promising direction to further elucidate motor degeneration along the full extent of the neuroaxis.

Conclusions

Sub-clinical central brain abnormalities are a potential feature of disease in SBMA associated with severity of peripheral neuropathy. The longitudinal profile of white matter abnormality in SBMA has yet to be reported but may offer insights into the trajectory of mutant AR disease activity, preferably from the pre-clinical stage. Total Neuropathy Score clinical version demonstrated differences in severity of peripheral neuropathy in patients and might benefit from prospective comparison of clinical utility or stratification for therapeutic trials given the absence of harmonized clinical screening tools for SBMA. Further studies are necessary to validate current findings.