Neurol India Home 

Year : 2021  |  Volume : 69  |  Issue : 4  |  Page : 889--893

A Study of Diffusion Tensor Imaging in Hirayama Disease

Jayantee Kalita1, Sushil K Rahi1, Sunil Kumar2, Suprava Naik2, Sanjeev K Bhoi1, Usha K Misra1,  
1 Department of Neurology, Sanjay Gandhi Postgraduate Institute of Medical Science, Lucknow, Uttar Pradesh, India
2 Department of Radio Diagnosis, Sanjay Gandhi Postgraduate Institute of Medical Science, Lucknow, Uttar Pradesh, India

Correspondence Address:
Prof. Usha K Misra
Department of Neurology, Vivekanand Poly Clinic and Institute of Medical Sciences, Aliganj, Lucknow - 226 007, Uttar Pradesh


Background: Hirayama disease (HD) is a motor neuron disease and occasionally is associated with lower limb hyper-reflexia. Corticospinal tract dysfunction can be evaluated by diffusion tensor imaging (DTI), but there is paucity of study in HD. Objective: We report corticospinal tract functions using DTI in the patients with HD and correlate with clinical findings. Materials and Methods: The patients with HD diagnosed on the basis of clinical and electromyography findings were included. Their age, duration of illness, side of initial involvement, and progression were noted. Presence of lower limb hyper-reflexia, and cervical spine magnetic resonance imaging (MRI) findings were noted. Cranial MRI was done and DTI findings at internal capsule, cerebral peduncle, pons, and pyramid were noted. Results: In total, 10 patients with HD and 5 matched controls were evaluated. The apparent diffusion coefficient (7.03 ± 0.27 vs 6.83 ± 0.36), fractional anisotropy (0.79 ± 0.04 vs 0.82 ± 0.05), axial diffusivity (5.08 ± 0.08 vs 5.04 ± 0.07), and radial diffusivity (3.79 ± 0.05 vs 3.76 ± 0.05) between HD patients and controls were not different in internal capsule. These values were also not significantly different in cerebral peduncle, pons, and pyramid. These values were also not significantly different between the severe and less severely affected sides. The fractional anisotropy did not correlate with lower limb hyper-reflexia (P = 1.00) and spinal cord atrophy (P = 0.60). Conclusion: DTI study in HD patients did not reveal corticospinal tract involvement in brain.

How to cite this article:
Kalita J, Rahi SK, Kumar S, Naik S, Bhoi SK, Misra UK. A Study of Diffusion Tensor Imaging in Hirayama Disease.Neurol India 2021;69:889-893

How to cite this URL:
Kalita J, Rahi SK, Kumar S, Naik S, Bhoi SK, Misra UK. A Study of Diffusion Tensor Imaging in Hirayama Disease. Neurol India [serial online] 2021 [cited 2022 Jan 24 ];69:889-893
Available from:

Full Text

Hirayama disease (HD) is characterized by unilateral or asymmetric oblique amyotrophy affecting C7-T1 myotomes.[1] HD is prevalent in Japan, India, Sri Lanka, Singapore, and Malaysia, and has been reported even from some European countries.[2] The exact etiology of HD is not well understood. The studies on SMN (survival motor neuron) gene deletion and CAG repeat sequence have failed to demonstrate genetic abnormalities.[3],[4] Various etiologies such as autoimmune, atopy, and spinal cord ischemia due to neck flexion have been suggested.[5],[6],[7] Reduction in amplitude of N13 of median somatosensory evoked potential during neck flexion has been reported and attributed to spinal cord compression or microvascular changes.[8] A subsequent study, however, did not confirm the changes of F wave parameters in neutral and neck flexion. There was also no change in latency or amplitude of N13 of median and ulnar somatosensory evoked potential in neutral and neck flexion.[9] Cervical spine magnetic resonance imaging (MRI) studies have revealed forward displacement of dural sac and flattening of lower cervical cord.[10] Passive dilatation of epidural venous plexus has been suggested on the basis of cinematographic MRI.[11] Some patients of HD have lower limb hyper-reflexia, which may be due to flexion myelopathy. Evaluation of soleus H reflex excitability and vibratory and reciprocal inhibition did not suggest involvement of corticospinal tract.[12] Diffusion tensor imaging (DTI) with tractography is a promising technique for demonstrating the microstructural changes especially white matter tracts in spinal cord and brain. There is paucity of DTI study evaluating the corticospinal tract in HD.[13] In this study, we report the corticospinal DTI changes in brain and correlate these with clinical findings in the patients with HD.

 Materials and Methods

The patients with HD attending neurology service of a tertiary care teaching institute were included. The diagnosis was based on the following criteria.[1]

Onset of disease between 15 and 25 years of age.Unilateral or asymmetric insidious amyotrophy often associated with cold paresisPolyminimyoclonus of fingers on moderate extension or presence of fasciculation on examination.Absent sensory, ocular, or sphincter disturbances.Little or no pyramidal signs.Arrest of disease after few years of onset.Neurogenic electromyography (EMG) changes restricted to C7-T1 myotomes.


Patients with history of stroke, head injury, perinatal hypoxia, seizure, mental retardation, and any other central nervous system disorders were excluded.


A detailed medical history including age of onset, duration of illness, and initial side of involvement, time to involve the other side, occupation, neck manipulation, cold paresis, fasciculation and polyminimyoclonus were noted. Presence of Horner's syndrome was noted. The topography and severity of muscle wasting, muscle power, tone, and tendon reflexes were recorded. Sensation of pinprick, temperature, joint position, and vibration were noted.


Blood counts, erythrocyte sedimentation rate for the first hour, hemoglobin, and serum chemistry were done. Needle EMG of abductor pollicis brevis, abductor digiti minimi, first dorsal interosseus, extensor digitorum communis, biceps, brachioradialis, triceps and deltoid on both sides and tibialis anterior, vastus lateralis, and gastrocnemius on the right side was done. Presence of spontaneous activity and morphology, duration, and interference pattern of motor unit potentials were noted.

Magnetic resonance imaging

All the patients underwent MRI on a 3T MR scanner (Signa Hdxt, GE, Milwaukee, WI, USA). Cervical MRI was done in neutral and flexed position of neck. T1 and T2 weighted MRI were obtained in axial and sagittal planes in neutral position. Postcontrast T1 sequence with and without fat suppression was obtained in flexion. DTI of brain was performed using SE/EPI sequence with the following parameters: TR (repetition time) – 6,000 ms, TE (echo time) – 85.5 ms, slice thickness 5 mm, and inter slice gap 1 mm, matrix 128 × 128, NEX – 1, band width – 250 KHz, field of view (FOV) – 28 × 28, number of directions – 25, and b-value – 1,000. DTI data were transferred to work station (ADW, GEMS, USA) for analysis. Postprocessing was done in functool diffusion tensor application. Region of interest (ROI) were placed in internal capsule, cerebral peduncles, pons, and medullary pyramids on both sides. Apparent diffusion coefficient (ADC), fractional anisotropy (FA), maximum, middle, and minimum Eigen values were calculated using the software supplied by the vendor (GE). Axial diffusivity (AD) is the maximum Eigen value. Radial diffusivity (RD) is the average of middle and minimum Eigen values. Mean diffusivity (MD) was taken as the average of maximum, middle, and minimum Eigen values.

Statistical analysis

The DTI results of the patients and the controls were compared by independent t-test. In the patients, the more severely affected limb was compared with unaffected or less severely affected limb by independent t-test. Analysis of variance was applied to compare DT1 findings in controls, severely affected, and less affected sides. These comparisons were done in different ROI. The DTI findings were correlated with age, duration, severity of wasting, cord atrophy on MRI, and enlargement of dorsal sac during flexion using Spearman or Karl–Pearson correlation test. A variable having a two-tailed P value of <0.05 was considered significant. The statistical analysis was done using SPSS 15 version and Graph Pad prism 5.


In total, 10 patients with HD and 5 controls were evaluated in this study. The age of the patients and controls was not significantly different (24.70 ± 5.77 vs 29.20 ± 2.59 years; P = 0.32). The median duration of illness was 2.5 (1–5) years. In all the patients, the disease started unilaterally; from left side in six and from right side in four patients. In five patients, the contralateral upper limb was also affected. Tone was normal but triceps reflex was reduced on the affected upper limb, although biceps and supinator reflexes were normal. Lower limb hyper-reflexia was present in six patients which was bilateral. The details are given [Table 1].{Table 1}

Concentric needle EMG revealed fasciculation and occasional fibrillations with long-duration, high-amplitude poly-phasic motor unit potentials in abductor digiti minimus, abductor pollicis brevis, extensor digitorum communis, and triceps on both sides. More severely affected side had poorer interference pattern. EMG of biceps, brachioradialis, deltoid, tibialis anterior, and vastus lateralis was normal. These finding was suggestive of chronic neurogenic changes in C7, C8, and T1 myotomes.

MRI study

Spinal MRI revealed lower cervical cord atrophy in three patients. There was enlargement of dural sac during neck flexion in four, which resulted thecal compression in three patients.

Cranial MRI did not reveal any abnormality in T1, T2, and FLAIR sequence. On DTI, the ADC value of internal capsule was 7.03 ± 0.27, cerebral peduncle 7.61 ± 0.59, pons 7.08 ± 0.92, and medullary zpyramid 7.98 ± 1.84. These values were not different between patients and controls. The FA, AD, and RD were also not different between patients and controls [Figure 1]. The details are shown in the [Table 2].{Figure 1}{Table 2}

The comparison of ADC, FA, AD, and RD values between severely affected and mildly affected sides did not reveal significant difference at different ROI [Figure 2]. None of the patients had FA value >1 and <0.1. The AD value of internal capsule of the severely affected side in the HD patients was significantly higher compared with the controls (P = 0.04). The remaining DTI parameters of the severely and less affected sides were not significantly different compared to controls.{Figure 2}

The FA value of internal capsule, cerebral peduncle, pons, and pyramid did not have significant correlation with duration of illness, lower limb hyper-reflexia, polyminimyoclonus, CMAP amplitude of median and ulnar nerve and MRI evidence of cervical cord atrophy, enlargement of posterior dural sac, and thecal compression. The details are summarized in [Table 3].{Table 3}


In this study, MRI revealed lower cervical cord atrophy in three patients with HD, but their DTI study did not reveal significant changes in FA, RD, AD, and ADC values at the level of internal capsule, cerebral peduncle, pons, and pyramid compared with the controls. The DTI values were also not different between severely and mildly affected sides. The FA values in these regions did not correlate with age of the patients, duration of illness, lower limb hyper-reflexia, CMAP amplitudes of ulnar and median nerves, and cervical cord atrophy. These results are consistent with sparing of corticospinal tract in the brain. Our results are in agreement with a study on seven patients with HD in whom DTI and central motor conduction time were evaluated. There was no difference in ADC, FA, AD, and RD in internal capsule, pons, and medulla compared with controls. There was also no difference on these DTI values between the first affected and later affected sides of HD patients and their central motor conduction time was also normal.[13] In this study, we have not evaluated central motor conduction time but we have reported central motor conduction time to upper and lower limbs in an earlier study in seven HD patients. The central motor conduction time to tibialis anterior was normal in all but to abductor digiti minimi was marginally prolonged in one side only. The spinal motor evoked potential to upper limb was delayed in two patients (two sides) and was attributed to the fall out of anterior horn cells. Two patients although had lower limb hyper reflexia but soleus H reflex parameters, such as HM ratio, vibratory inhibition, and reciprocal inhibition were normal suggesting normal corticospinal tract function in the lower limbs.[12] In a case study on HD, dynamic DTI study of spinal cord revealed reduced FA values in left posterior and anterior hemicord during neutral and neck flexion position.[14] DTI study of spinal cord is challenging and is limited by susceptibility artifacts; hence, we have not done DTI of cervical spinal cord.

In a volumetric analysis of corticospinal tract in amyotrophic lateral sclerosis patients using DTI, there was significant reduction in corticospinal tract volume. There was, however, no correlation between corticospinal tract volume and the clinical parameters.[15] The primary application of tractography has been mainly done for visualization of white matter trajectories in 3D. The water diffusion in tissues is highly sensitive of differences in microstructural architecture of cellular membrane. The apparent diffusivity is increased, if there is increase in the average spacing between the membrane layers, and ADC value decreases, if the space is reduced. These properties render DT1 technique as a powerful tool for detecting microscopic differences in tissues. In normal white matter, FA value range between 0.1 and 1.0 and the variation can occur due to crossing of white matter fibers. FA value is decreased where there is least crossing of white matter. The FA value is likely to decrease in corticospinal tract involvement and in the areas where there is haphazard arborization.

In our patients, EMG revealed chronic neurogenic changes in C7-T1 myotomes in all the patients on either side. The studies from China, however, have reported neurogenic changed beyond C7-T1 myotomes on EMG.[16],[17] In classical HD, neurogenic changes, however, should restrict to C7-T1 myotomes. Almost exclusive occurrence of HD in males, familial occurrence and normal corticospinal tract functions on DT1, central motor conduction time, and somatosensory evoked potential studies reinforce the concept that HD is a segmental spinal muscular atrophy and it may have a genetic basis.

Ethical statement

This study involves human participants in study.


We thank Mr. Shakti Kumar for secretarial help.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Hirayama K. Non-progressive juvenile spinal muscular atrophy of the distal upper limb (Hirayama's disease). In De Jong JM, editor. Handbook of Clinical Neurology. Vol 15. Amsterdam, The Netherlands: Elsevier; 1991, p. 107-20.
2Hirayama K. Juvenile muscular atrophy of distal upper extremity (Hirayama disease). Intern Med 2000;39:283-90.
3Misra UK, Kalita J, Mishra VN, Kesari A, Mittal B. A clinical, magnetic resonance imaging, and survival motor neuron gene deletion study of Hirayama disease. Arch Neurol 2005;62:120-3.
4Kalita J, Misra UK, Mishra DK, Thangaraj K, Mittal RD, Mittal BR. Non-progressive juvenile-onset spinal muscular atrophy: A clinico-radiological and CAG repeat study of androgen receptor gene. J Neurol Sci 2007;252:24-8.
5Yoon JH, Joo IS, Li WY, Sohn SY. Clinical and laboratory characteristics of atopic myelitis: Korean experience. J Neurol Sci 2009;285:154-8.
6Kira J, Ochi H. Juvenile muscular atrophy of the distal upper limb (Hirayama disease) associated with atopy. J Neurol Neurosurg Psychiatry 2001;70:798-801.
7Schröder R, Keller E, Flacke S, Schmidt S, Pohl C, Klockgether T, et al. MRI findings in Hirayama's disease: Flexion-induced cervical myelopathy or intrinsic motor neuron disease? J Neurol 1999;246:1069-74.
8Restuccia D, Rubino M, Valeriani M, Mirabella M, Sabatelli M, Tonali P. Cervical cord dysfunction during neck flexion in Hirayama's disease. Neurology 2003;60:1980-3.
9Misra UK, Kalita J, Mishra VN, Phadke RV, Hadique A. Effect of neck flexion on F wave, somatosensory evoked potentials, and magnetic resonance imaging in Hirayama disease. J Neurol Neurosurg Psychiatry 2006;77:695-8.
10Hirayama K, Tokumaru Y. Cervical dural sac and spinal cord in juvenile muscular atrophy of distal upper extremity. Neurology 2000;54:1922-6.
11Tokumaru Y, Hirayama K. Anterior shift of posterior lower cervical dura mater in patients with juvenile muscular atrophy of unilateral upper extremity. Rinsho Shinkeigaku 1989;29:1237-43.
12Misra UK, Kalita J. Central motor conduction in Hirayama disease. Electroencephalogr Clin Neurophysiol 1995;97:73-6.
13Boelmans K, Kaufmann J, Schmelzer S, Vielhaber S, Kornhuber M, Münchau A, et al. Hirayama disease is a pure spinal motor neuron disorder-a combined DTI and transcranial magnetic stimulation study. J Neurol 2013;260:540-8.
14Cunningham RK, Shah P, Evan C, Barakat N, O'Connor E, Luo JJ. 3D MRI-diagnostic, dynamic and diffusion tensor evaluation of Hirayama disease. Neurological cases 2014;1:4-7.
15Wang S, Poptani H, Bilello M, Wu X, Woo JH, Elman LB, et al. Diffusion tensor imaging in amyotrophic lateral sclerosis: Volumetric analysis of the corticospinal tract. Am J Neuroradiol 2006;27:1234-8.
16Wang XN, Cui LY, Liu MS, Guan YZ, Li BH, DU H. A clinical neurophysiology study of Hirayama disease. Chin Med J (Engl) 2012;125:1115-20.
17Guo XM, Qin XY, Huang C. Neuroelectrophysiological characteristics of Hirayama disease: Report of 14 cases. Chin Med J (Engl) 2012;125:2440-3.