Discussion
SMA-LED and CMT2O are related to DYNC1H1 mutations and are characterized by predominant wasting and weakness of lower limbs, delayed motor milestones, and abnormalities on electromyography (EMG) examination. In this study, we reported two CMT2O cases with heterozygous DYNC1H1 mutations c.1792C>T/p.R598C and c.790C>G/p.R264G, respectively. The former mutation of DYNC1H1 is, to our knowledge, the first one reported in East Asia that causes CMT 2O, and the latter is a novel dominant mutation reported for the first time worldwide. In addition, we analyzed the clinical phenotypes of neuromuscular diseases caused by DYNC1H1 mutations and relevant genotype-phenotype correlation in previous publications. We compared the clinical characteristics between the two disease entities. We found that all CMT2O patients were presented with distal-dominant wasting, while proximal-dominant lower limb wasting can be observed in some cases of SMA-LEDs. We also found that patients with mutations in the DYN1 region may have more severe phenotype than those mutated in the DHC_N1 region.
Our results showed that weakness and wasting of lower limbs were the most common clinical phenotype of DYNC1H1- related neuromuscular diseases. Delayed motor milestones were also observed in a large percentage of patients. The most common phenotype on physical examination was decreased or absent tendon reflex. This was in accordance with previous animal experiments that the Swlheterozygous mutation of Dync1h1 led to absent H reflex in mice, even if the motor nerve function was normal (Chen et al., 2007).
The most common clinical features of SMA-LED and CMT2O, including delayed motor milestones and lower limb-predominant wasting and weakness, are often not specific, especially in the early stage of the diseases. Typically, SMA-LED and CMT2O are diagnosed and differentiated according to EMG results. On EMG examination, SMA-LED is characterized by motor axonal or motor neuron loss, which is in the form of active denervation, compensatory reinnervation and enlargement of the motor unit action potential (Arnold, Kassar, & Kissel, 2015). CMT2O usually presents a typical pattern of axonal neuropathy affecting both motor and sensory modalities on EMG, similar to other CMT type2 diseases (Pakhrin et al., 2018). Our study also suggested the NCV test as another applicable electrophysiological approach to differentiate these disease entities, as abnormal NCV result may be more commonly seen in SMA-LED than that in CMT2O. However, in the early stage of these diseases, electrophysiological abnormalities may not be typical enough to make an accurate diagnosis. Therefore, considering the common variants of gene that may lead to SMA-LED and CMT2O, clinical phenotypes may be an essential characterization in the early diagnosis and differentiation of these two disease entities. In our study, it was found that all CMT2O patients were presented with distal-dominant wasting, while proximal-dominant lower limb wasting can be observed in some cases of SMA-LEDs. Besides, SMA-LED patients had a higher percentage of proximal-dominant weakness of lower limbs, while CMT2O patients more frequently presented distal-dominant weakness. This may suggest that although lower limb involvement is a common feature of both diseases, they may have different distribution pattern of muscle impairment, which may provide some insights for clinicians when making an early diagnosis. In addition, abnormal sensation was more common in CMT2O patients, consistent with earlier reports that SMA-LED patients are usually exempt from sensory impairment (Arnold et al., 2015). Therefore, the involvement of the sensory system may be a relatively strong indicator of CMT2O in patients with DYNC1H1 mutation presenting relevant clinical manifestations.
DYNC1H1 mutations involved in our study were located either in the DHC_N1 region or in the DYN1 region. Not only did patients with mutations in DYN1 present with more severe atrophy and weakness of lower limbs, but they may also show clinical phenotypes more complicated and difficult to make a right diagnosis. Besides, manifestations of CNS impairment, such as epilepsy, cognitive impairment, ADHD and abnormal brain MRI was more common in patients with mutations in the DYN1 region. This is in accordance with earlier finding that mutations in the tail domain (close to the N-terminal) usually lead to pure motor neuron deficits hardly with any signs of brain abnormalities (Harms et al., 2012); (Rossor et al., 2012), while mutations in the motor domain (close to the C-terminal) often cause malformation of cortical development (MCD) and epilepsy (Poirier et al., 2013). Interestingly, no DYNC1H1 mutation related to motor neuron diseases has been observed located in the C-terminal (Dynein_C region). This suggests a highly clinical heterogeneity of DYNC1H1 variants, the mechanism of which warrants further investigation, such as epigenetic modification, environmental exposure, etc. In addition, as precision medicine is becoming more and more advocated, this genotype-phenotype correlation ofDYNC1H1 mutation may serve as a reference for accurate diagnosis and personalized treatment.
Our study also suggested that some mutation sites of DYNC1H1 had a higher recurrence rate than other sites, of which exon 8 seemed to be the mutational hot region and c.1792C>T/p.R598C was the most frequent point mutation. However, no c.1792C>T/p.R598C mutation has been found in the East Asian population at present, although some other recurrent DYNC1H1 mutations have been detected in East Asia. On one hand, this discrepancy may reflect differences on genetic level among different races; on the other hand, this result may be due to the relatively limited amount of cases in this area. Considering the large population basis in East Asia, more different kinds of DYNC1H1 mutations may be gradually detected, as our team recently identified a teenage patient with CMT2O caused by c.1792C>T/p.R598C heterozygous mutation in DYNC1H1 . Up to now, no specific treatment for SMA-LED and CMT2O has been put forward yet, thus symptomatic treatment and orthopedic surgery remain important in order to decrease the risk of complications of these diseases as early as possible. Therefore, genetic testing is of necessity in early diagnosis and prompt treatment. Besides, gene treatment, such as genome editing, provides new insight into treating diseases caused by pathogenic gene mutations, especially those with high recurrent rate. In this regard, early genetic testing for clinically suspected SMA-LED or CMT2O patients will be illustrating for applying appropriate therapeutics and trying new therapeutic techniques.
The conclusions of this study shouldn’t be assessed without the consideration of several limitations. Firstly, not all patients have a well-recorded phenotype, though the literature were systematically reviewed. Thus, the study did not present a complete mutation spectrum since patients without any detailed description of clinical features are not studied. Secondly, the reporting bias of included articles cannot be ignored because cases with recurrent variants and classical phenotypes were less likely to be reported. Thirdly, all these comparisons are based on cases in the literature and online database. Therefore, the genotype-phenotype correlations should be validated in an independent cohort. Despite the limitations above, our study provided a comprehensive summary of the DYNC1H1 mutations spectrum by selecting SMA-LED and CMT2O diseases and unrevealed the hidden correlation between genotype and phenotype.