Selective Vulnerability of Nerve Demyelination in Chronic Acquired Multifocal Polyneuropathy

The pathological features of peripheral nerve myelin lesion are known as diffuse and focal demyelination. As a hallmark of peripheral nerve focal demyelination, conduction block (CB) has been found in a series of both acquired and hereditary demyelinating neuropathies [1-4]. Multifocal motor neuropathy (MMN) and Lewis-Sumner syndrome (LSS) are the most representative chronic acquired focal neuropathies, while many of the classic chronic inflammatory demyelinating polyradiculoneuropathy Received: June 06, 2019

(CIDP) also presents with CB [5]. Distinct from common peripheral neuropathies, the motor impairments in these focal neuropathies are clinically uneven, i.e., non-length dependent and asymmetric [6,7]. Their limb onset selectivity is well-known [8], however, not efficient in reciprocal differentiation. Different nerve involvement among demyelinating polyneuropathies was proposed by the morphological study, which suggested the selective vulnerability of nerves [9]. The selectivity was observed in other neuropathies [10,11], but hasn't been fully studied in focal demyelination by electrophysiology methods. Thus, in this study, by using of a bunch of well-established electrophysiological indicators, e.g., distal motor latency (DML), terminal latency index (TLI), F-wave latency and nerve conduction velocity (NCV) as well as CB, we aimed to investigate the selective vulnerability and the demyelinating pattern of nerves and segments in patients with MMN, LSS and CIDP with CB (CIDP-CB).  [13]. The exclusion criteria were peripheral neuropathy family history; alcohol abuse; toxic and neurotoxic exposure; tumor; metabolic disorders including pathoglycemia (diabetic mellitus and impaired glucose tolerance); paraproteinemia (monoclonal gamma-globin related polyneuropathy).

Clinical Profiles and Examinations
General information of gender, age and height were recorded.
Detailed clinical history was investigated to collect disease duration and onset and determine neurological manifestations of weakness, numbness and atrophy. Meticulous physical examinations were performed to determine tendon reflex as well as other positive neurological signs. Blood routine test, liver and kidney function, blood electrolytes, thyroid function, vitamin B 12 , folic acid and blood glucose were checked to meet the inclusion and exclusion criteria.
Gangliosidosis antibodies examination were semi-quantitative by blot and was fulfilled by KingMed Diagnostics (Guangdong, China). A panel of various kinds of anti-Gangliosidosis (GM1, GM2, GM3, GD1a, GD1b, GQ1b, GD1b) was used and positive result was determined by visible band in GM1 area. Blood immunofixation electrophoresis was tested for subjects to meet the exclusion terms.
Lumbar puncture was performed under the informed consent of subjects and cerebrospinal fluid (CSF) was quantitatively analyzed to determine albumin protein concentration.

Nerve Conduction
All subjects received nerve conduction test by Keypoint.
net (Natus, CA, USA) in a quiet room with temperature >20℃.
Surface electrodes were used to record wave form on fully relaxed muscle after stimulation. Median, ulnar, radial, peroneal, peroneal superficial, tibial and sural nerve were selected for evaluation (sensory nerve data were not shown). Multiple sites were stimulated for every nerve to obtain sufficient data and calculate velocities. For  Normal values and ranges of our lab were adopted. Negative peak amplitude, curve area and duration were recorded to determine CB and temporal dispersion (TD). A definite CB was defined as area reduction on proximal vs. distal stimulation of at least 50%, distal compound muscle action potential (CMAP) must be >20% of the lower limit of normal and >1 mV, and increase of proximal to distal CMAP duration must be ≤30%.13 Nerve variant especially Martin-Gruber anastomosis was carefully excluded by additional stimulation. Temporal dispersion was defined as >30% duration increase between the proximal and distal CMAP 12 .
Take-off latency and distance were recorded and NCV was thus calculated. Distal distance (dD), DML, and NCV was used to calculate TLI by the following formula: TLI=(dD/DML)/NCV. F-wave was performed during a bout of stimulation recorded on distal muscle and the first F-wave latency (F) was recorded. DML, proximal motor latency (PML) which is the latency recorded from the second stimulation site from distal, and F were used to calculate modified F-wave ratio (MFR) by the formula:

Statistical Analysis
Median, ulnar, radial, peroneal and tibial nerves were selected for studying CB and TD distributions. Median, ulnar, peroneal and tibial nerves were selected for comparing other conduction parameters. Normal distribution data were described as mean ( x − ) ± standard deviation (SD), and skewed distribution data were described as median (25% percentile -75% percentile).

CB And TD Distribution of Nerves
CB was more likely to occur around elbow (p<0.01) while less likely to be found between wrist and elbow in CIDP-CB (p<0.05).
CB was found in more than half of tibial nerve examined in MMN CB distribution is shown in Table 2.   Table 3.

Discussion
MMN and LSS were characterized by male predominant, upper limb onset and different sensory involvement [14][15][16][17]. In this study we also found a male predominant composition in the MMN group and LSS group and it was different in the CIDP-CB group. High upper limb onset proportion and rare sensory involvement in MMN were also proved and could help in differentiating CIDP-CB. In addition, we found MMN presented slowly progressive long duration, low CSF protein and frequent positive anti-GM1, while LSS and CIDP presented progressive duration, moderate-high CSF protein and rare positive anti-GM1. Clinical findings were confirmatory and consistent with previous study [18,19]. Using electrophysiological indicators, we presented a horizontal comparison of LSS, MMN and CIDP-CB nerves. Selective vulnerability of nerve is particularly interesting and controversial in focal neuropathy. Ulnar nerve has been suggested useful in diagnosing acute demyelinating neuropathy [20], and tibial nerve was hypothesized to be especially involved in MMN [21], while other study found no difference in CB distribution [22]. In this study, noting that wrist to elbow segment of ulnar nerve conduction was least likely to be blocked in CIDP- The mechanism of nerve selectivity of chronic focal demyelinating diseases hasn't been fully investigated. Recently, a common etiology discovered in demyelinating diseases with CB was the detection of autoantibodies, e.g., anti-neurofascin 140/186, anti-neurofascin 155 [23,24]. As a target, the neurofascin expression in node and paranode sites determines the impairment pattern of conduction failure and is associated with CB [25]. A reasonably hypothesis is that the uneven distribution of nerve expression of neurofascin may contribute to the nerve selectivity.
Distal vulnerability was suggested in typical CIDP previously [26,27], while, CIDP-CB, as a subtype, has rarely been analyzed and compared solely. Although prolonged DML in all CIDP-CB nerves was in accordance with the distal predominant pattern, the similar values of TLI, a comparison of distal and middle segments [28], of most nerves among groups implied that the distal impairments might commensurate with middle segment when the existence of CB was taken into consideration. The CB also masked the different diffuse demyelinating of median nerve as only ulnar nerve MCV was lower in CIDP-CB. When the nerves with CB were excluded, upper limb MCV decrease in CIDP-CB emerged. In the conduction study, we proved that CIDP-CB showed many common features with typical CIDP comparing to MMN and LSS and this highlighted its diffused demyelination feature in median and ulnar nerves.

Conclusion
We reported the difference of selective segmental vulnerability of ulnar and tibial nerves in LSS, MMN and CIDP-CB, and proved more diffuse demyelinating features in CIDP-CB which distinguished it from LSS and MMN. The different nerve demyelination distribution and pattern could help in differentiation. This study was limited by insufficient proximal and distal conduction data and small sample.
A prospective cohort and immunological mechanism study are anticipated.