Benign monomelic amyotrophy of lower limb in a cohort of chinese patients

Abstract Background Benign monomelic amyotrophy of lower limb (BMALL) is a neurogenic syndrome representing an unclear field. Further studies might be helpful to elucidate uncertainties regarding causation, outcome, and the risk of progression to amyotrophic lateral sclerosis (ALS). Methods According to the inclusion and exclusion criteria, 37 patients with BMALL were retrospectively collected in three neuromuscular centers from January 2012 to October 2018. The detailed medical data were summarized. Multiple laboratory tests were examined. Routine electrophysiological examinations, muscle MRI of lower limbs, and muscle biopsy were conducted. Results The cohort included 24 male and 13 female cases with median age of onset 47 years. Muscle MRI revealed that the distribution of involved muscles matched with the extent of fat infiltration, so the pattern muscle atrophy can be divided into the following four types: six patients with thigh atrophy (type I), 14 patients with leg atrophy (type II); 10 patients with disproportionate atrophy in both thigh and leg (type III); and seven patients with well‐proportionate atrophy in both thigh and leg (type IV). Electrophysiological findings showed neurogenic pattern, spontaneous activity, and abnormal H reflex, which suggested a disorder of spinal anterior horn cell in the patients with types I‐III. However, no electrophysiological abnormalities were found in the patients with type IV. Muscle pathology varied from almost normal pattern to advanced neurogenic pattern in nine biopsied patients. Follow‐up showed that two patients with type II developed to ALS four years later, and all patients with type IV were in stable condition without any complaints. Conclusion Muscle MRI was useful to exactly localize the distribution of involved muscles in BMALL patients. The distribution of atrophic muscles can be roughly divided into four types based on the MRI features. The classification of distributing types might be as an indicator for the prognosis of BMALL.


| INTRODUC TI ON
Benign monomelic amyotrophy of lower limb (BMALL) is a rare neurological syndrome (Felice et al., 2003;Gourie-Devi, 2007). It is clinically characterized by insidious-onset wasting restricted to a single lower limb with a slow progressive or arrested course, although a few deteriorating cases have occasionally been described (De Carvalho & Swash, 2007;Kay et al., 1994). The cause of BMALL and its relation to other anterior horn cell disorders, such as amyotrophic lateral sclerosis (ALS), are poorly understood (Guennoc et al., 2013).
The majority of BMALL cases were found in male Indian (Gourie-Devi et al., 1984;Nalini et al., 2014;Prabhakar et al., 1981;Saha et al., 1997;Virmani & Mohan, 1985), although some cases were also reported in East Asia (Hamano et al., 1999;Kim et al., 1994) and Western countries (Freitas & Nascimento, 2000;Visser et al., 1988;Muzio et al., 1994;Dimachkie et al., 2000;Guglielmo et al., 1996;Moglia et al., 2011;Münchau & Rosenkranz, 2000;Riggs et al., 1984;Uncini et al., 1992). Only a few cases were reported in Mainland China up to now (Hui et al., 2018), although China has the largest population in the world. BMALL represents an unclear field among neurogenic syndromes; further studies might be helpful to elucidate uncertainties regarding causation, outcome, and the risk of progression to ALS. In this study, we aimed to describe the clinical characteristics, electrophysiological changes, MRI features, and histopathological patterns in a cohort of 37 Chinese patients with BMALL.

| Subjects
All patients with monomelic amyotrophy of lower limbs were col-

| Ethical statement
All patients' tissue samples were obtained after a written consent signed by each individual in compliance with the bioethics laws of China and the Declaration of Helsinki. The research was approved by ethics committee of the First Affiliated Hospital of Nanchang University.

| Muscle MRI
Axial planes of the thigh and leg muscles were imaged in all patients using 3.0-T MR scanners. Conventional T1-weighted image (T1WI) sequences were obtained to observe fatty infiltration with the following parameters: repetition time (TR) = 500ms, echo time (TE) = 8 ms, and matrix: 512 × 512. The extent of fatty infiltration was evaluated according to the modified Mercuri scale (0-5 scores) (Mercuri et al., 2002). A fatty score was adopted as follows: normal (score 0); punctate hyperintense (score 1); fatty streaks detected below 30% muscle volume (score 2); hyperintense detected among 30% to 60% muscle volume (score 3); hyperintense detected above 60% muscle volume (score 4); and hyperintense in the whole muscle (score 5). The short time inversion recovery (STIR) sequences were obtained to evaluate muscle edema with the following parameters: TR = 6100ms, TE = 70ms, inversion time = 180ms, and matrix: 512 × 512. The slice thickness was 5mm with a slice gap of 1mm, and the field of view was 36 × 48cm.

| Muscle biopsy
Muscle biopsies were conducted in nine patients. The tissues were frozen and then cut into 8µm sections. These sections were stained according to standard histological and enzyme histochemical procedures with hematoxylin and eosin (H&E), modified Gomori trichrome, periodic acidic Schiff, oil red O, adenosine triphosphatase (ATPase), nicotinamide adenine dinucleotide tetrazolium reductase, succinate dehydrogenase, cytochrome c oxidase, and nonspecific esterase. Immunohistochemical stain was applied with myosin heavy chain 7 antibody (MYH7, Abcam) to indicate the type 1 fibers.

| Statistical analysis
Data were analyzed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA Differences were considered statistically significant if p <.05.

| Patient enrollment
A total of 49 patients were initially recruited based on the inclusion criteria, but 12 of them were excluded due to the following reasons: Five

| Clinical features
The cohort included 24 male and 13 female patients without family history. The median age of first hospital visiting was 51 (44, 56) years. The median age of onset was 47 (38, 53) years. The occupation of the patients included labor worker (17 cases), student (9 cases), athlete (3 cases), staff member (3 cases), and unemployment (5 cases). Nine patients had no symptoms until their limb wasting was occasionally noticed by themselves or others, so the definite age of onset was difficult to determine in some patients.
All patients exhibited muscle atrophy of a single lower limb, but 11 patients simultaneously had complaint of muscle weakness at the first visiting. Eighteen patients had concomitant symptoms including cold paresis in seven patients, muscle fasciculation in four patients, subjective numbness in four patients, muscle soreness in three patients, local skin pigmentation in two patients, and myalgia in one patient. The main clinical characteristics of patients are summarized in Table 1.
The values of erythrocyte sedimentation rate, hemoglobin, blood sugar, thyroid function hormone, parathyroid hormone, and serum immunoglobulins were within normal limits. The level of CK was elevated in five patients with 234, 340, 294, 304, and 750 IU/L, respectively, and others had a range from 51 to 210 IU/L (normal 30-210 IU/L). Serum IgM of antipoliovirus was negative in 12 examined patients. Serum IgM of anti-GM1 antibody was negative in 11 examined patients. CSF analysis of 10 patients showed that the level of glucose and chloride all were normal; the white blood cell counts with normal range from 0 to 8 x 10 6 /L in eight patients, the other two patients were 9 and 12 x 10 6 /L, respectively; the level of protein with normal range from 150 to 450 mg/L in nine patients, the other one was 520 mg/L.
Based on the distribution of involved muscles from a clinical view, we identified four types of muscle atrophy ( Figure 1): restricted to thigh atrophy (type I); restricted to leg atrophy (type II); disproportionate atrophy in both thigh and leg (type III); and wellproportionate atrophy in both thigh and leg (type IV). The type I group included 6 patients (3 males and 3 females). The median age of onset was 44 years. The wasting distribution showed anterior thigh atrophy in four patients and posterior thigh atrophy in two patients. The type II group included 14 patients (10 males and 4 females). The median age of onset was 45 years. The wasting distribution showed whole leg atrophy in seven patient, posterior leg atrophy in six patients, and anterior leg atrophy in one patient.
The type III group included 10 patients (7 males and 3 females).
The median age of onset was 33 years. The wasting distribution showed posterior thigh and posterior leg atrophy in four patients, posterior thigh and anterior leg atrophy in four patients, and anterior thigh and posterior leg atrophy in two patients. The type IV group included seven patients (4 males and 3 females). The median age of onset was 12 years. The wasting distribution showed a wellproportionate atrophy of the whole lower limb in comparison with contralateral lower limb.
The patients with type IV had a younger age of onset with comparison to those of the other three patterns (Table 2). Linear regression model was configured to further evaluate the confounders that may have introduced bias into the retrospective study. After adjustment for sex, duration of illness, occupation, muscle weakness, and

| Electrophysiological changes
In the patients with type I wasting, the median motor nerve conduc- reflex was not evoked in six patients, the latency was prolonged in the 3 patients, and the latency was normal in one patient.
In the patients with type IV wasting, the motor and sensory nerve conduction studies of the wasting limbs were normal. The H reflexes were normal in examined patients.

F I G U R E 1
The distribution of involved muscles can be divided into four types. Type I was restricted to thigh atrophy represented by case 1 with fat infiltration in quadriceps femoris. Type II was restricted to leg atrophy represented by case 10 with diffused fat infiltration in the leg. Type III was disproportionate atrophy in both thigh and leg represented by case 24 with a combination fat infiltration of thigh posterior muscle group and leg anterior muscle group. Type IV was well-proportionate atrophy in both thigh and leg represented by case 31 without fat infiltration. The patients gave their written informed consent for the publication The results of needle EMG are listed in supplemental Table S1 in detail. Overall, most clinically involved muscles of lower limbs showed chronic reinnervative changes in patients with type I, type II, and type III, while the muscles of upper limbs were intact. The

| Muscle MRI features
Muscle MRI revealed that the distribution of involved muscles ex-  Table S2 and Table S3.

| Muscle pathological changes
The muscle pathological changes were variable in the nine patients underwent muscle biopsy. Two patients displayed almost normal pattern except for a few angular atrophy fibers (Figure 2a). One patient showed a mild neurogenic muscle atrophy characterized by groups of relatively small atrophy in some foci (Figure 2b). Three patients exhibited a typical neurogenic muscle atrophy characterized by a large group of angular atrophy in a whole muscle fascicle ( Figure 2c). Three patients appeared an advanced neurogenic muscle pattern featured by groups of severely atrophic fibers with pyknotic nuclear clumps and chains, muscle hypertrophy, proliferation of connective tissue, internal migration of nuclei, and occasionally degenerative fibers (Figure 2d). The neurogenic groupings were further confirmed by MHY7 immunostaining for type 1 fiber ( Figure 2e) and ATPase staining at PH 10.6 for type 2 fiber (Figure 2f).

| D ISCUSS I ON
Unilateral atrophy of a lower limb could be noticed by the patients because of wasting with or without weakness on walking, but in over a half of the patients it was incidentally observed by their family members, friends, or physicians during consultation for unrelated illness (Gourie-Devi, 2007;Nalini et al., 2014). In our study, only 11 patients initially complained of mild weakness of unilateral lower limb, and the others displayed insidious wasting of lower limb.
Consequently, the precise age of onset and illness duration may not be very accurate. The age of onset of our patients was similar to that of European (Carvalho & Swash, 2007) or Japanese population (Hamano et al., 1999), but elder than that of Indian (Nalini et al., 2014) or Korean population (Kim et al., 1994). Remarkable gender preference with a male/female ratio 2:1 to 40:1 had been reported in BMALL patients from different population groups [12,24]. In our case series, the male/female ratio was 1.7:1, which was similar to that of Korean population (Kim et al., 1994), and indicating that female patients might be relatively common in Eastern Asia (Table 3). The patients with type IV had a younger age of onset and a relatively equivalent sex ratio with comparison to those of other three types, which suggested that the condition of type IV might be a distinct subgroup.
Neurogenic pattern of EMG, spontaneous activity, and abnormal H reflex indicated a disorder of spinal anterior horn cell in our patients.  but no muscle edema was observed in all patients, which suggested that the lesions were in chronic and stable processes. According to the different distribution of involved muscle groups, we divided the pattern of muscle atrophy into four types. More than half of the patients had fat infiltration in the posterior calf muscles, which had lower frequency than that in Indian and Western patients (Dimachkie et al., 2000;Hamano et al., 1999). MRI muscle examination is very useful to detect the involved muscles, especially deep muscles, and help us locate the affected segments of spinal cord. For example, MRI showed the selective involvement of adductor magnus and tibialis anterior muscles in the patient 10 ( Figure 1, type III); the adductor magnus was innervated by a branch of sciatic nerve from L4 to L5, and the anterior tibialis was innervated by peroneal nerve from L4-S1; therefore, the neuronal lesion should be at anterior sciatic nerve root from L4 to L5 or anterior horn of L4 toL5 spinal cord (Uncini et al., 1992). More intriguingly, MRI showed a well-proportionate atrophy of both thigh and leg muscles without fat infiltration in the seven patients with type IV, which indicated that the atrophy process might be due to a congenital development disorder.
The muscle pathology varied from almost normal to advanced neurogenic pattern in our patients. The large grouping and muscle fiber hypertrophy suggested that the original lesion might be located at spinal anterior horn cells. The myopathic-like changes seen in some of the sections may be secondary to occur in ALS and various neurogenic atrophies (Chen et al., 2018). Therefore, the significance of muscle biopsy was limited to patients with BMALL, because the detailed electrophysiological examination and conventional muscle MRI can offer enough diagnostic evidences for BMALL. It was a limitation not to obtain the muscle biopsy from the patients with type IV, which might be helpful to explore the underlying pathogenesis.
BMALL is a rare condition characterized by wasting of a single lower limb with a stable course. Though labeled as "benign" because of a nonprogressive course, a few deteriorating cases were also reported in BMALL patients (Carvalho & Swash, 2007). Similarly, two patients developed to ALS at the 39th month and 56th month after initial evaluation in our follow-up. However, it was difficult to determine whether these patients with BMALL could progress to ALS at a late stage because about 10% of the patients with ALS still appeared as pure motor neuron involvement in the lower limb 6 years after the disease onset (Berg-Vos et al., 2003). Interestingly, we found that patients with ALS progression were all in the type II group, though the statistical analysis showed a borderline but negative significance between different groups. Future study with a large number of cases might elucidate the definite relation between clinical outcome and classification of subgroup.
The cause of BMALL is still uncertain, though monomelic amyotrophy of the lower limb has a definite neurogenic origin. BMALL has several similitudines with Hirayama disease (HD). Both syndromes are common in the Asian countries, asymmetrically involve a single limb (lower limb in BMALL and upper limb in HD), and have an insidious onset and a self-limiting course. Therefore, the BMALL might have a similar pathogenesis with HD that is wildly reported in more cases and intensively investigated in the pathogenic aspects (Foster et al., 2015).
This study had some limitations that need to be explicitly acknowledged. First, it was a retrospective study; thus, some clinical data were incomplete that would lower the clinical significance, for example, the panel of ganglioside antibodies and CSF examinations were not conducted in all patients; therefore, it was difficult to exclude the possibility of immune-related motor neuropathy in some patients. Similarly, the post-polio syndrome could not completely be excluded in some old patients, though no history of poliomyelitis and negative IgM antibody refuted the condition (Abrar & Ahmad, 2015).
Second, the EMG results were retrospectively collected in these patients, so the examined muscles were not normalized in each patient, which undermined the reliability of diagnostic protocols.
Additionally, EMG data were collected from 3 different hospitals, the normal values of the electrophysiological test were a little different in different clinics, so the examined values were just expressed as medians and quartile interval, and not emphasized the absolute values and change range. Third, the follow-up time was not enough to differentiate BMALL from ALS in some patients. Commonly, a follow-up > 3-4 years was considered a useful criterion to exclude the majority of patients with ALS (Berg-Vos et al., 2003;Weiss, 2005).
Although 28 of 37 patients had more than five-year follow-up plus disease duration, it was reasonable to follow-up BMALL patients as long as possible.
In summary, it was conclusive that the monomelic amyotrophy of the lower limb had a neurogenic origin, though it was a syndrome comprising a group of conditions with different etiology. Muscle MRI was useful to exactly localize the distribution of involved muscles.
The muscle atrophy can roughly be divided into four types on the basis of the distribution of fat infiltration on MRI. Patients with type II showed a tendency toward ALS, but patients with type IV maintained a clinical stabilization without any risks of ALS.

ACK N OWLED G M ENTS
We thank the patients and their families for cooperation. We also thank Ms. Yaqing Yu for pathological technician.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R CO NTR I B UTI O N
WL and XH contributed to analysis, interpretation, and drafting. CS, WH, ZY, CR, and WH contributed to the acquisition and analysis of data. LJ and JK performed the pathological study. ZM and ZMH performed the electrophysiological analysis. YS and HD contributed the study design and drafting the manuscript. HD contributed funding acquisition.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/brb3.2073.

DATA AVA I L A B I L I T Y S TAT E M E N T
All relevant data are within the paper and its Supporting Information files.