Introduction

Sialidosis is an autosomal recessive disease with pathogenic variants in the NEU1 gene [1]. It is a rare disease with an estimated prevalence of 1 in 4 million births [2]. Sialidosis encompasses different symptoms, including myoclonus, seizures, ataxia, and visual disturbance, with variated onset ages [3, 4]. Myoclonus is induced by various minor stimulations, such as sound stimuli, light touch, and passive and voluntary movements. Cherry-red spots in the macula are often [4] – but not universally [5,6,7] –observed in patients, along with degeneration of the retina.

The pathogenesis has been proposed to be reduced neuraminidase activity leading to sialyloligosaccharide accumulation in affected tissues [1]. The reduced neuraminidase activities result in oversialylated lysosomal membrane protein 1, which in turn results in excessive exocytosis of lysosome contents. Pathologically, neurons in the central nervous system may show excessive sialyloligosaccharide accumulation in the cytoplasm. Radiologically, diffuse cerebral atrophy may be observed, with the cerebellar region prone to involvement, although this finding is not universal.

Involvement of the central nervous system (CNS) in sialidosis, presented as hyperexcitability and disinhibition, had been well studied [8, 9]. However, the involvement of the peripheral nervous system (PNS) has rarely been reported. Here we report the clinical and electrophysiological features in the PNS among patients with sialidosis.

Results

We recruited six nonconsanguineous patients with genetically-confirmed sialidosis from unrelated Taiwanese families; two of the patients are siblings (Case 5 &6). Four patients were female. The mean age at initial disease presentation was 10.3 ± 5.2 years. The clinical phenotypes, genotype, enzyme activity assays, and electrophysiological findings are summarized in Tables 1 and 2, and Table 3. Among the typical symptoms of sialidosis, five patients had myoclonus at the time of evaluation, while ataxia or cerebellar signs were noted in four patients. Hyperreflexia and spasticity were observed in all patients. Four patients had cherry red spots. Among the four patients who received brain MRI, only one patient presented with mildly delayed myelination near both the frontoparietal central parenchyma and both posterior internal capsules. Four patients (Case 1 since the age of 21 years old, Case 2 since the age of 18, Case 3 since the age of 24, Case 6 since the age of 21) were wheelchair-bound, and the other two patients still walk independently on the examinations. (Case 4: 18 years old, Case 5: 33 years old).

Table 1 Clinical profiles of patients with sialidosis
Full size table
Table 2 Electrophysiological studies in patients with sialidosis
Full size table
Table 3 Findings from nerve conduction studies in patients with sialidosis*
Full size table

Notably, four patients experienced painful paresthesia, and three patients reported it as the initial presentation. The symptoms were often induced by heat or by exercise and subsided after resting or removal of the inducing factors. None of the patients reported significant limb deformities. There was no observable atrophy, fasciculation, or significant sensory loss.

Electrophysiological studies of the PNS

The motor and sensory nerve conduction velocities were in the normal range in patients with sialidosis. Case 3 had a reduced CMAP amplitude in the right ulnar nerve, while case 5 had prolonged distal motor latency in the bilateral tibial and peroneal nerves. There were no repetitive CMAP observed. Five patients (83.3%) had abnormalities in the late response study, all of whom had prolonged F-wave latency, especially in the lower limbs (median nerve: 27.39 ± 1.27, 25.5–30.0 ms; ulnar nerve: 26.50 ± 1.91, 22.3–28.9 ms; peroneal nerve: 55.18 ± 9.00, 46.4–74.1 ms; tibial nerve: 54.76 ± 5.48, 46.2–62.2 ms). Repeater F-waves were noted in 3 patients. Three patients accepted the EMG study, and two patients had neurogenic polyphasic waves without spontaneous activities in sampled muscles. Sympathetic sensory response studies, done in 2 patients, revealed normal results. Only Case 3 received quantitative sensory testing, and the result is normal.

A unique very late response other than F-wave was noted

Beyond the typical F-waves, we also identified a special very late response in all of the recruited patients in the F-wave study. These responses have a latency that is much longer than the upper limit of F-wave latency in our lab (50 ms). The amplitude of these late responses ranged from 5 mV to 40 mV, and could present as either monophasic or multiphasic. The latency of the very late responses ranged from 70.8 to 176.4 ms. These very late responses could coexist with the F-wave (such as in patient 3, Fig. 1A) or appear when the F-wave was absent (Fig. 1B). This prolonged late response is most commonly presented in the tibial nerves, but could also be identified in the median and ulnar nerves.

Fig. 1
figure 1

Findings from electrophysiological studies in patients with sialidosis. (A) In the F-wave study, very late responses (arrow) with a latency of 74 ms could be observed compared to the normal F waves with a latency of 42 ms (arrowhead). (B) These very late responses (latency: 70.8 ms) may exist without the presence of the F-wave

Full size image

Discussion

Sialidosis is considered to be a disease with primary involvement of the CNS, and the involvement of PNS was rare reported. Our study revealed two intriguing findings: (1) the peculiar phenomenon in the late response study and (2) the painful paresthesia at the early stage of disease.

An intriguing finding in our study is the abnormalities in late response study, including the changes in the persistence and latency of F-wave, the presence of a repeater F-wave, and a “very late response” with high amplitude. The change in F-wave latency with normal motor conduction velocity is considered to correspond to a lesion in the proximal peripheral nerve or spinal cord. The repeater F-waves have been proposed to be related to lower motor neuron dysfunction or disinhibition due to upper motor neuron dysfunction [9,10,11]. The involvement of upper and lower motor neurons in previous autopsy studies supported that the change in F-waves was related to sialidosis [11, 12].

The presence of very late responses and giant waves has not been reported in previous studies. The difference in the shape and amplitude with normal F-waves, in addition to having a significantly prolonged latency, indicated that its pathology may be differ from traditional changes in F-waves [13, 14]. The proximal conduction time of the F wave could be calculated as 0.5*(F-M-1), with F being the latency of the F-wave, M being the distal latency, and the central conduction time estimated to be 1 millisecond. A very prolonged F-wave could be observed in Charcot-Marie-Tooth disease type 1, because the motor nerve conduction velocity is very slow. In the very late response observed in our patients, the latency was significantly prolonged compared with the estimated F-wave latency, which was calculated from the distal latency (M), motor nerve conduction velocity (MCV), and estimated distance between the stimulation and cord (D): \(Estimated F \left(ms\right) =M+\frac{2D}{MCV}+1.\) Since the nerve conduction velocity and distal latency remained largely within normal limits, a prolonged central conduction time (much longer than 1 ms) is likely to be the cause of the changes. On the other hand, the origin of enlarged F-wave amplitude, previous studies had reported a possible relationship with lower motor dysfunction [15] although the involvement of the upper neurons could not be ruled out definitely [16]. Study on the myoclonus generator of sialidosis also revealed evidence indicating involvement of subcortical circuits, which result in motor hyperexcitability [17]. Considering that interneuron and upper motor neuron modify the excitability of lower motor neurons, we assume that the prolonged latency may result from aberrant interneuron cycles within the spinal cord, while the increased amplitude may be caused by disinhibition from the cortical-subcortical pathways, as had been reported by previous studies of sialidosis.

Our study revealed that patients may present with intermittent painful paresthesia as an initial symptom of sialidosis, which has rarely been described in previous studies. A case series documented that one of the 5 patients reported having painful neuropathy [18], while sensory deficits were found among 8 of 17 patients in another study [19]. In our study, all patients reporting these symptoms did not have continuous numbness or other negative symptoms. In the evaluation, all patients reported normal results on clinical sensory examination and quantitative sensory testing. As a result, this symptom should belong to positive symptoms, which may be related to ectopic activities or central disinhibition. In a previous pathological study, storage materials were noted in the Schwann cells of sural nerves [8]. Sialic acid accumulations had also been reported to be present in the dorsal root ganglion [12]. Combining the electrophysiological study of the peripheral nerve system that we provided, the origin of painful paresthesia may be the peripheral nerve. However, the hyperexcitability described above may also contribute to patients’ presence of painful paresthesia, for which the development of hyperalgesia and allodynia may be due to excessive responses and inadequate suppressive mechanisms to aberrant ectopic activities or benign sensory stimulations experienced by the patient [17].

Our study has several limitations. First, despite active recruitment of study subjects, we could only enroll 6 cases in our study. This is largely due to the rare nature of sialidosis, and study of a larger cohort with longer follow-up may reveal more abnormalities in the PNS for this disease. Second, many other techniques had been developed to further evaluate the pathophysiology of peripheral nerve disorders, such as nerve/skin biopsy, peripheral nerve imaging, and nerve excitability study. Incorporation of these techniques in future studies may lead to better understanding of the pathophysiology of this unique phenomenon in sialidosis.

Conclusion

This study observed a novel very late response in the nerve conduction study of patients with sialidosis. This finding suggested that patients with sialidosis may present with involvement of the PNS, which is parallel with CNS involvement. The very late response may indicate abnormalities in the inhibition pathway in patients with sialidosis, warranting further studies.

Methods

Patients and clinical profiles

This retrospective study recruited patients with genetically confirmed sialidosis at National Taiwan University Hospital. All patients had detailed evaluations, including history taking, neurological examinations, image study, and electrophysiology studies, including electroencephalogram, somatosensory evoked potentials, visual evoked potentials, nerve conduction studies, F-wave studies, and electromyogram.

Nerve conduction studies, F-wave studies, and electromyogram

Nerve conduction studies and electromyograms were performed using a Viking IV Electromyographer (Nicolet, Madison, WI) in all patients following established methods. The bilateral median, ulnar, tibial, peroneal, and sural nerves were studied. The results of nerve conduction studies were compared to normative data in our laboratory.

For the evaluation of the F-wave, the bilateral median, ulnar, tibial, and peroneal nerves were studied using the same equipment. Ten to 15 stimulations were applied at a frequency of 1 Hz. The minimal latency and persistence of F-waves were recorded. The minimal latency of the F wave was defined as the shortest latency of the presented F-wave as identified during the study. The persistence of the F-wave was defined as the percentage of definable F-waves in all stimulations. Repeater F waves were defined as F-waves with the same shape, latency, and amplitude as identified by the examiner. Electromyography was performed by a board-certified neurologist who determined the site for evaluation and interpretation of the results.

Statistical analysis

Numerical variables are expressed as the mean ± SD. The minimal and maximal values were also included. All analyses were performed using Stata software (StataCorp LP, College Station, TX).