Characteristics of Hirayama Disease in Young South Korean Soldiers

Hirayama disease (HD), also known as benign focal amyotrophy, monomelic amyotrophy, and juvenile benign muscular amyotrophy, is a neurologic disorder that was initially described by Hirayama in Japan in 1959.1 HD is diagnosed based on a combination of distinctive clinical features and electrodiagnostic and cervical magnetic resonance imaging (MRI) findings.2, 3, 4, 5, 6 The clinical presentation of HD includes asymmetric oblique amyotrophy, specifically involving the distal muscles of the hand and forearm, but sparing the brachioradialis. HD primarily affects males and typically manifests during late adolescence or early adulthood.5, 6 The disease progresses slowly over a period of 3–5 years before reaching a self-limiting phase.4, 5, 6, 7 HD is more common in Asian populations,8 and while some South Korean case reports have been published, there have been no extensive population studies.9, 10, 11, 12, 13, 14, 15, 16

South Korea implements compulsory military service, and all South Korean males normally undergo a physical examination for military enlistment when they reach 19 years old. Males who are deemed unfit for military service, such as those with HD, are exempted from the obligation to perform military service at the Military Manpower Administration (MMA). However, in cases where a diagnosis has not been made prior to military enlistment or symptoms develop after military enlistment, the individuals are diagnosed at a military hospital while serving and are discharged from military service by the review committee. A previous Medical Command and MMA data study including 192 South Korean males with HD in their early 20s over a 10-year period found that 157 (81.8%) were exempted from military service before military enlistment and 35 (18.2%) were exempted during military service.17 Although it is difficult to access the specific medical records of patients with HD who have been exempted from military service by the MMA, it is meaningful to investigate the clinical presentations in patients with HD diagnosed at a military hospital. These data could serve as representative HD samples of the young male population in South Korea.

The objective of this study was to determine the clinical features, electromyography (EMG) findings, and MRI characteristics of HD in young South Korean males through a retrospective review of medical records using data from young South Korean soldiers diagnosed with HD during their military service over a 10-year period.

Patients diagnosed with HD during their military service and discharged by the review committee from 2011 to 2021 were identified from Medical Command data. Only enlisted soldiers with ranks of private, private first class, corporal, and sergeant were included, with officers being excluded.

Medical records were reviewed retrospectively to collect demographics, chief complaints, clinical presentations, the presence of bilateral or unilateral symptom manifestations, initial dates of outpatient visits to military hospitals, the initial departments where patients were evaluated, the onset of the symptoms, family history, and the results of EMG and cervical MRI.

We reviewed the medical records comprehensively to identify the presence of typical and atypical clinical manifestations in HD. We focused on assessing the following clinical parameters: weakness and muscle atrophy of upper extremities, occurrence of cold paresis, sensory disturbances such as paresthesia or hypesthesia, and the existence of pyramidal signs.

The EMG findings were analyzed to identify neurogenic changes such as denervation or reinnervation potentials within the sampled muscles and to determine the proportion of each muscle involvement. We also investigated whether the muscles exhibiting neurogenic changes were unilaterally or bilaterally affected and whether these changes were consistent with the reported clinical symptoms.

The imaging analysis was conducted by two physicians according to the protocol reported by Chen et al.18 Flexion-position MRI was used to investigate the presence of anterior displacement of the dural sac resulting from a crescent-shaped mass in the posterior epidural space. Neutral-position MRI was used to evaluate the presence of loss of attachment (LOA) between the posterior dura and subjacent lamina, asymmetric cord flattening, loss of lordosis, localized cord atrophy, and noncompressive intramedullary T2-weighted high signal intensity. We also evaluated the level at which cord atrophy was most severe. In cases without neutral-position MRI, loss of lordosis was evaluated using neutral cervical X-rays, and localized cord atrophy and noncompressive intramedullary T2-weighted high signal intensity were assessed using flexion-position MRI.

Forty-two male patients aged 20.2±0.8 years (mean±standard deviation) were identified. The most common complaint was hand weakness (100%), followed by hand tremor in 23.8% (Table 1). Four patients (9.5%) complained of asymmetric symptoms bilaterally in the upper extremities, 27 (71.1%) had symptoms only on the right side, and 11 (28.9%) had symptoms only on the left side. Sensory disturbance was evaluated in patients with symptoms of hypesthesia or paresthesia, and this was found in five patients (11.9%). Pyramidal signs were not observed in any patient.

Table 1
Clinical characteristics of the included patients with Hirayama disease

• Click for larger image

Table 2 presents the onset durations of HD in the patients according to their military ranks at the time of discharge from military service and the timing of symptom onset (before or after enlistment). The patients included 25 individuals (59.5%) who developed symptoms prior to military enlistment, with an onset duration of 20.6±12.3 months. In contrast, 17 patients (40.5%) experienced symptoms after their military enlistment, with an onset duration of 4.6±2.3 months.

Table 2
Comparison of onset duration of HD according to symptoms before and after enlistment by military rank

• Click for larger image

EMG demonstrated that muscles in the C8–T1 myotome were the most commonly involved, including the flexor carpi ulnaris (89.7%), first dorsal interosseous (88.6%), abductor digiti minimi (88.5%), and abductor pollicis brevis (74.3%) (Table 3). Additionally, most muscles in the C7 myotome were commonly involved, including the extensor digitorum communis (78.9%), extensor indicis proprius (73.7%), flexor carpi radialis (70.4%), and triceps (25.9%). Some cases showed involvement of muscles in the C5–C6 myotome, such as the biceps brachii (12.5%) and deltoid (10.5%), but not the brachioradialis.

Table 3
Electrodiagnostic findings in patients with HD

• Click for larger image

Four (17.4%) of the 23 patients who underwent needle EMG in both upper extremities had symptoms in both upper extremities. Needle EMG revealed neurogenic changes on the contralateral side in 8 (42.1%) of the 19 patients who had symptoms in a single upper extremity (Table 3).

A crescent-shaped mass was observed in the posterior epidural space in 25 (83.3%) of the 30 patients who underwent flexion-position cervical MRI. Cervical MRI revealed that localized cord atrophy (90.0%) was the most-prominent finding, with atrophy being most severe at the C5–C6 disc level (58.3%), followed by at the C6–C7 disc level (25.0%) and the C6 vertebral body level (11.1%) (Table 4). Loss of lordosis and the LOA sign were found in 81.8% and 59.4% of patients, respectively, while asymmetric cord flattening (37.5%) and noncompressive intramedullary T2-weighted high signal intensity (10.0%) were less common in MRI (Fig. 1).

Fig. 1
Representative MRI findings in HD patients. A: Localized cord atrophy at the C5–C6 level. B: Loss of lordosis. C: Crescent-shaped mass in the posterior epidural space. D: Asymmetric cord flattening. E: Loss-of-attachment sign. HD, Hirayama disease; MRI, magnetic resonance imaging.

• Click for larger image

Table 4
Cervical imaging findings in patients with HD

• Click for larger image

This study analyzed clinical symptoms and EMG and MRI results of 42 young South Korean male patients with HD, representing the largest reported sample of South Korean patients with HD to date. The lack of large-scale studies on HD in South Korea makes the present analysis of 42 patients diagnosed with HD especially important.

Kim et al.17 noted that HD was the second most common neuromuscular disease after myasthenia gravis in soldiers. The high prevalence of HD in the military is attributed to the age at disease onset. Many studies have found that HD symptoms begin between the ages of 17 and 19 years, with most patients experiencing symptoms during their late teens to early 20s.4, 6, 19 In our study the patients were aged 20.2±0.8 years and had an onset duration of 20.6±12.3 months. These findings were consistent with the results of previous studies on symptom onset.

While 17 patients (40.5%) experienced symptom onset after military enlistment, it is highly unlikely that the onset of HD actually occurred after enlistment given that the diagnosis was made at 4.6±2.3 months after enlistment and that EMG confirmed the presence of chronic denervation. However, it is possible that the symptoms exacerbated during military service. Most of the subjects in our study showed the insidious aggravation of symptoms, but three individuals experienced symptom exacerbation following specific training activities such as cold-weather training, marksmanship training, and extended marches in full military gear.

Typical clinical manifestations of HD include weakness and atrophy in the distal upper extremities, an insidious onset and progression over 3–5 years, tremors in the affected hand, worsening of symptoms in cold environments (referred to as cold paresis), and the absence of sensory symtpoms.1, 4, 5, 6, 19 All of the patients in our study exhibited distal hand weakness and atrophy, with tremors observed in 23.8% of patients. Previous studies found that the prevalence of cold paresis ranged from 27.3% to 97%,5, 6 where it was only 4.7% in our study. This much smaller proportion of patients with cold paresis is thought to have been due to the inherent limitations of a retrospective review of medical records.

Several studies found that patients with HD exhibit atypical clinical features with pyramidal signs, involvement of the proximal upper extremities, and sensory disturbances.3, 6, 11, 12, 19, 20 Pyramidal signs, which refer to upper motor neuron signs, have been reported in 2.4% of Japanese and 10.6% of Chinese patients with HD, predominantly in severe cases.6, 19 None of our subjects showed pyramidal signs. Sensory disturbances were reported in 19.2% of 333 HD patients in Japan and in 30% of 40 HD patients in Taiwan.19, 21 In our study, five of the patients with HD (11.9%) showed associated sensory disturbances.

Skin discoloration related to HD has not been reported previously, but this was present in two individuals in the present study. It is plausible to attribute this symptom to changes in blood flow due to hand weakness and autonomic nervous system abnormalities. According to Gourie-Devi and Nalini,22 the sympathetic nervous system is implicated not only in the atrophic upper limb of HD patients but also in the unaffected upper limb. The occurrence of skin color changes only in some cases suggests that they can be attributed to hand weakness and autonomic nervous system abnormalities rather than to the disease itself.

Most patients with HD present with distal upper extremity symptoms. However, several studies have found proximal upper extremity involvement in some patients with HD.3, 12, 19, 20, 23 Huang et al.21 documented proximal involvement in 3 of 40 patients (7.5%), while Gourie-Devi and Nalini24 observed proximal involvement in 4 of 44 patients (2.3%). In our study, none of the 42 patients complained of symptoms involving the proximal upper extremities, but 4 patients (9.5%) showed denervation or reinnervation potentials in the deltoid and biceps muscles in EMG. According to Wang et al.,3 proximal HD often involves a more-extensive range of cervical spinal segments, affecting both rostral and caudal areas around C5–C7, whereas the simple distal type primarily affects the C7–T1 myotome. Wang et al.3 also reported that proximal HD often presents with a higher incidence of cervical kyphosis. In our study, all four patients with proximal muscle involvement in EMG exhibited kyphosis in MRI, and three of them displayed extensive cord atrophy, including at the C5–C7 levels, which is consistent with previous findings. The presence of these features in proximal HD suggests that this variant of the disease is more severe than simple distal HD.

HD is characterized by asymmetrically unilateral amyotrophy, with some patients experiencing bilateral symptoms.4, 6 In our study, 90.5% of patients complained of symptoms unilaterally in the upper extremities and the remaining 9.5% complained of asymmetric symptoms bilaterally in the upper extremities; none of the patients complained of symmetric bilateral symptoms. Among those with unilateral symptoms, 71.1% reported symptoms on the right side and 28.9% reported symptoms on the left upper limb. Most research studies have found that the right upper limb is affected more often than the left upper limb, at a ratio of approximately 3:1,4, 7, 21 which is consistent with our findings. The asymmetric involvement of the upper extremities is thought to result from uneven flattening of the cervical spinal cord. Various hypotheses have been proposed to explain this asymmetry.4, 6, 7 The most-popular hypothesis is that the posterior dural sac shifts forward during neck flexion, resulting in uneven compression of the lower cervical cord and causing ischemic damage to motor neurons. Another hypothesis is that an uneven distribution of posterior epidural ligaments contributes to asymmetric cord compression. While a correlation between right-handedness and symptoms predominantly affecting the right side might be expected, previous studies have found no significant association with handedness.7, 19 The reason for the greater impact on the right side than the left remains unclear, indicating the need for further research.

Eight (42.1%) of the 19 patients who had symptoms in the unilateral extremity and underwent needle EMG in both extremities showed neurogenic changes in the contralateral limb. Nalini et al.20 analyzed 279 cases over a 35-year period and reported neurogenic changes in the contralateral limb in 20% of patients with unilateral limb symptoms, which is consistent with the findings of other studies.

Neck-flexion MRI is the primary diagnostic modality for HD.18, 25, 26, 27 The obtained images reveal a characteristic crescent-shaped, well-enhanced mass in the posterior epidural space that results from anterior displacement of the tight dural sac and widening of the posterior epidural space.19, 28 However, several studies have found that not all cases of HD show a crescent-shaped mass in the posterior epidural space in neck-flexion imaging.29, 30 Indeed, we observed a crescent-shaped mass in the posterior epidural space in only 25 of the 30 patients who underwent flexion-position MRI.

In addition to neck-flexion MRI, neutral-position MRI provides valuable clues for the diagnosis of HD.18 Several studies have obtained abnormal results when applying neutral-position MRI to patients with HD. According to a comparison between 46 patients with HD and 51 controls performed by Chen et al.,18 the LOA sign was the most common finding (identified in 93% of patients with HD), while asymmetric cord flattening and localized lower cervical atrophy were observed in 70% and 59%, respectively. We found that localized cord atrophy was the most common finding (90.0%), with the LOA sign being less common (59.4%). However, similar to our results, several other studies have found cord atrophy to be common in patients with HD. Paladi et al.27 identified cord atrophy in 80% of 15 patients, and Sonwalkar et al.31 reported that all patients showed localized cord atrophy while 50% showed the LOA sign.

Our assessments of cervical cord atrophy in cervical MRI revealed that it was most severe at the C5–C6 disc level (58.3%, n=21), followed by at the C6–C7 disc level (25.0%, n=9), which is consistent with previous literature. Several studies have found varying degrees of cord atrophy above and below the C6 vertebral body level.2, 32 Some studies have found atrophy to be most severe at the C6–C7 level,25, 29 with others finding a higher prevalence of C5–C6 cord atrophy.27

Our study had several limitations that warrant discussion. First, this study only included male conscripts in South Korea, and so HD was not investigated in females. HD exhibits a male predominance, but the reported male-to-female ratio varies in the literature.6 Given that HD primarily occurs in males, our study comprising only male participants still provides valuable information about HD in the South Korean population. Second, we excluded early-onset cases due to the inclusion criterion of military conscripts. Among the male soldiers with HD aged in their early 20s in South Korea, 81.8% are reportedly diagnosed before military enlistment and 18.2% during military service. We investigated patients with HD who were diagnosed at military hospitals, which therefore did not include individuals who developed the disease at a lower age. Third, the retrospective design of this study meant that not all of the data were available for every case. The low prevalence of symptoms such as cold paresis might have resulted from underreporting and EMG not being routinely conducted on the contralateral side.

In conclusion, this is the first large-scale study to analyze the clinical characteristics and electrodiagnostic and imaging studies of young male patients with HD in South Korea. The clinical presentation and features of electrodiagnostic and imaging studies revealed in this research will help to improve the understanding of HD in the young South Korean male population.