Criteria for conduction block based on computer simulation studies of nerve conduction with human data obtained in the forearm segment of the median nerve
1 Department of Clinical Neurophysiology, Neuromuscular Research Group, Rudolf Magnus Institute of Neuroscience, University Medical Centre Utrecht The Netherlands 2 Department of Neurology, Neuromuscular Research Group, Rudolf Magnus Institute of Neuroscience, University Medical Centre Utrecht The Netherlands.
Correspondence to: H. Franssen, MD, PhD, Department of Clinical Neurophysiology, University Medical Centre Utrecht, PO Box 85500, 3508 GA Utrecht, The Netherlands E-mail: h.franssen{at}umcutrecht.nl
 |
Summary |
|---|
 Top Summary IntroductionPatients and methodsResultsDiscussionReferences |
|---|
The finding of conduction block (CB) on nerve conduction studies supports the diagnosis of potentially treatable immune-mediated neuropathies. CB in a number of axons may result in reduction of the compound muscle action potential (CMAP) on proximal versus distal stimulation (decrement). Decrement may also result from increased temporal dispersion (TD) as this leads to desynchronization and phase cancellation of the motor unit action potentials (MUAPs) out of which the CMAP is built up; polyphasia of MUAPs possibly yields additional decrement. To prove the occurrence of CB, decrement has to be larger than can be explained by increased TD or increased phase cancellation. This was established previously by simulations using MUAPs recorded in rats assuming maximal TD. Unfortunately, criteria based on human data and criteria for nerves with limited TD are not available. In the present study, criteria for CB were derived using simulations with thenar surface recorded MUAPs affected by collateral reinnervation that were obtained in patients with lower motor neurone disease (LMND). The effect of TD on decrement was determined for a wide range of TDs in the forearm segment of the median nerve and the segment distal to this. Our criteria for CB were based on area decrement because this was less influenced by TD and more by CB than amplitude decrement. The maximal area decrement in the forearm segment increased as TD in the forearm segment increased but decreased as TD in the distal segment increased. This suggests that, when desynchronization and phase cancellation occur in the distal segment due to TD, less phase cancellation and, therefore, less decrement can occur due to TD in the forearm. The finding that duration prolongation on proximal versus distal stimulation reflected TD within the forearm segment and that distal duration reflected TD in the distal segment allowed proposal of a more flexible set of criteria for forearm segments when TD in the forearm segment is limited or TD in the distal segment is pronounced. A separate investigation showed that the maximal TD in chronic inflammatory demyelinating polyneuropathy was within the range of our simulations, indicating that these were realistic. Our criteria were validated retrospectively in patients with multifocal motor neuropathy and patients with LMND. In the forearm segment of the median nerve, our criteria were more sensitive and equally specific for CB as compared with criteria for CB based on the study using rats. Our criteria have to be evaluated prospectively.
Key Words: conduction block; diagnosis; neuropathy
Abbreviations: CB, conduction block; CIDP, chronic inflammatory demyelinating polyneuropathy; CMAP, compound muscle action potential; LMND, lower motor neurone disease; MMN, multifocal motor neuropathy; MUAPs, motor unit action potentials; TD, temporal dispersion
Received October 29, 2005. Revised June 22, 2006. Accepted June 26, 2006.
|
Introduction |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
The finding of conduction block (CB) on nerve conduction studies supports the diagnosis of potentially treatable neuropathies that are characterized by immune-mediated demyelination (Dyck et al., 1982
; Lewis et al., 1982; Albers and Kelly, 1989; Cornblath, 1990; Van Doorn et al., 1990; AAN, 1991; Kaji et al., 1992; Chaudhry et al., 1993; Nobile-Orazio et al., 1993; Hahn et al., 1996a, b; American Association of Electrodiagnostic Medicine, 1999; Hughes et al., 2001; Dalakas, 2002; Berger et al., 2003; Olney et al., 2003). CB is defined as the failure of action potential propagation at a given site along a structurally intact axon (Kaji, 2003). CB in a number of axons of a nerve segment may result in a reduction of the compound muscle action potential (CMAP) on proximal versus distal stimulation (decrement). Abnormal decrement can also be the result of increased temporal dispersion (TD), which is an increase in the difference between the conduction times along the different axons within a nerve (Lee et al., 1975; Olney et al., 1987; Rhee et al., 1990; Oh et al., 1994). Abnormal decrement may arise because increased TD yields desynchronized arrival as well as increased cancellation between the positive and negative phases of the motor unit action potentials (MUAPs) out of which the CMAP is built up. Phase cancellation was suggested to be exaggerated when the MUAPs contributing to the CMAP are polyphasic due to collateral reinnervation following partial denervation (Cornblath et al., 1991; Brown and Bolton, 1993).
The current criteria for CB provide insufficient evidence that conduction is actually blocked when they are fulfilled. To prove the occurrence of CB, decrement has to be larger than can be explained by increased TD or increased phase cancellation. As CB, increased TD and collateral reinnervation may occur together in nerves of patients with immune-mediated neuropathies, comparative studies between these patients and control groups are not suitable for deriving criteria that are specific for CB (Brown and Feasby, 1984; Van den Berg-Vos et al., 2002; Van Asseldonk et al., 2003). Instead, computer simulation of nerve conduction is required to study the effects of the different mechanisms which may contribute to decrement. In a simulation study, which provided important insights, the effect of TD on decrement was determined (Rhee et al., 1990). CMAPs were reconstructed from MUAPs that were recorded by a subcutaneous needle electrode in healthy rats (Rhee et al., 1990). The results indicated that TD, which was either large or deliberately chosen to yield a maximum decrement, could result in a decrement of CMAP area of up to 50%. It was concluded that an area decrement of >50% could no longer be explained by TD and, therefore, indicated CB in at least some axons. Although this criterion is widely used, it is open to debate whether it is applicable in clinical practice for the following reasons. Most importantly, CMAPs were not reconstructed from human MUAPs. Furthermore, the MUAPs were recorded using subcutaneous needle electrodes, whereas CMAPs on nerve conduction studies consist of a summation of surface-recorded MUAPs. Also, the effect of polyphasic MUAPs on decrement was not taken into account. Finally, because this criterion assumes maximal effects of TD on decrement, it is likely to be insensitive for the detection of CB.
In the present study, criteria for CB were derived based on the maximal decrement that may result from increased TD and altered MUAP shape due to collateral reinnervation in the human median nerve. Our criteria assume CB when decrement is larger than can be explained by these factors. Our study consists of different parts. First, the effects of various amounts of TD on decrement were determined by computer simulation of median nerve conduction using surface-recorded human thenar MUAPs for CMAP reconstruction. To obtain MUAPs of which the shape was altered due to collateral reinnervation, these were recorded in patients with lower motor neurone disease (LMND). Second, we assessed whether a more flexible set of criteria for CB can be derived when TD is limited. For this purpose, we determined which CMAP variables, readily available from nerve conduction studies, reflect TD. Third, using the above described computer simulation, we blocked different percentages of axons in the nerve and assessed which CMAP variables were most sensitive for the detection of CB. Fourth, we assessed whether the simulated TD was realistic by estimating the maximal TD in patients with chronic inflammatory demyelinating polyneuropathy (CIDP). Finally, we validated our criteria by applying them retrospectively in patients with multifocal motor neuropathy (MMN) and patients with LMND.
|
Patients and methods |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Patients
To obtain polyphasic MUAPs recorded with surface electrodes, we investigated patients with LMND. This disorder is characterized by axonal degeneration and collateral sprouting resulting in polyphasic MUAPs on concentric needle EMG. We investigated the records of 49 patients and selected eight patients in whom at least 30% of MUAPs were polyphasic on previously performed concentric needle EMG of the m. abductor pollicis brevis. Four patients were diagnosed as segmental LMND, two as slowly progressive LMND and two as distal LMND. The mean age was 51 years (range 38–67) and six were men. In these eight patients, 12 muscles were examined.
To asses the maximal TD, which may occur as a result of demyelination in human nerves, we estimated TD in patients with CIDP because pronounced TD occurs in CIDP. We investigated the records of 50 patients with CIDP and selected five patients with the largest duration for the prolongation of the CMAP on proximal versus distal stimulation of the forearm segment of the median nerve. We estimated the latency difference between the fastest and slowest conducting axons in the median nerve by means of collision techniques. The median age of these five patients was 37 years (range 28–78) and all were men.
All participants gave informed consent to the study, which was approved by the Medical Ethics Committee of the University Medical Centre in Utrecht.
MUAP recording in patients with LMND
Surface MUAPs were recorded using the spike-triggered averaging technique (Stein and Yang, 1990). Surface AgAgCl recording electrodes with a diameter of 10 mm were placed on the m. abductor pollicis brevis in a belly tendon montage (bandpass 5 Hz to 10 kHz). The location of the grid 1 electrode was adjusted until supramaximal stimulation of the median nerve at the wrist yielded a CMAP with the highest amplitude. Intramuscular MUAPs were recorded by a monopolar tungsten needle electrode (bandpass 500 Hz to 10 kHz). During voluntary abduction of the thumb against resistance, the needle-recorded MUAPs were used to trigger the recording of surface MUAPs between 10 ms before and 40 ms after the beginning of the trigger signal. At a stable level of contraction, the surface EMG signals were averaged up to a maximum of 360 discharges until the surface MUAP could be clearly and reproducibly delineated from the baseline. Both low and medium threshold motor units were recruited by having the subject vary the force of muscle contraction. This method allowed recording of surface MUAPs up to moderately strong levels of contraction (Milner-Brown et al., 1973). In each muscle, three insertions were made on a line perpendicular to the length of the muscle fibres. With each insertion, MUAPs were sampled at increasing depths. For each muscle, the size, shape and firing frequency of averaged surface MUAPs were compared to ensure that a given surface MUAP was not included more than once. Surface MUAPs were digitized and transferred to an MS DOS computer for further analysis. For each surface MUAP, we determined amplitude, duration and area of each negative and positive phase. A polyphasic needle-recorded MUAP was defined as a MUAP with more than four baseline crossings (Brown and Bolton, 1993). A polyphasic surface MUAP was defined as an MUAP with more than one baseline crossing.
Prior to all investigations, including the collision studies (see below), the arm was warmed in water at 37°C for 45 min (Franssen and Wieneke, 1994). During investigation, the arm was kept at 37°C by infrared heaters and the fingers were fixated by a splint.
CMAP reconstruction
For each of the 12 muscles from which MUAPs were recorded, a simulated CMAP was obtained by arithmetical summation of all surface MUAPs recorded from that muscle (Biro and Partridge, 1971; Lee et al., 1975; Nandedkar and Stalberg, 1983; Rhee et al., 1990). Next, the number of surface MUAPs for that muscle was multiplied until the amplitude of the negative phase of the simulated CMAP was equal to that of the real CMAP recorded by stimulation of the median nerve at the wrist. Thus, one MUAP set for each of the 12 muscles was created. The recorded CMAP was first reconstructed by using a TD of zero (see below).
Simulation of TD in a forearm segment
To study the effect of TD on decrement in a nerve segment that represents the forearm segment of the median nerve on nerve conduction studies, a distal CMAP was simulated for a conduction distance of 7 cm to the muscle and a proximal CMAP was simulated for a conduction distance of 37 cm to the muscle. In our study, TD is defined as the difference in conduction time (latency) between the slowest conducting axon and the fastest conducting axon within a nerve. The length of the distal segment (7 cm) reflects the distance from the wrist to the grid 1 recording electrode on the m. abductor pollicis brevis. A forearm segment with a length of 30 cm was chosen because 95% of the patients within our patient population had a forearm nerve segment of 30 cm or less (data not shown).
To simulate a distal CMAP, each MUAP within a set was assigned a latency according to the following procedures. Within a MUAP set, the latency of the fastest MUAP was fixed at 1.17 ms (excluding the latency caused by neuromuscular transmission) corresponding to a conduction velocity of 60 m/s over 7 cm. First, distal segments without TD were simulated. This enabled us to study separately the effects of TD on decrement in the forearm segment. This was done because we showed that TD in the distal segment affected the decrement in the forearm segment that could occur owing to TD in the forearm segment (see results). Next, distal segments with TD were simulated to study the combined effects of TD in the distal and forearm segments on decrement in the forearm segment. To simulate distal segments with various amounts of TD, the latency of the slowest MUAP was varied in 20 steps of 
1 ms, so that it was 0–20 ms slower than the latency of the fastest MUAP (0 ms represents no TD; 20 ms represents maximal TD). The latencies of the remaining MUAPs were assigned according to four models. In model 1, latencies were assigned in an amplitude-dependent manner: the larger the amplitude of the negative MUAP phase the smaller the latency. In models 2, 3 and 4, latency assignment was random and independent of MUAP size. In model 2 more short than long latencies were assigned. In model 3 more long than short latencies were assigned. In model 4 all latencies were assigned with equal frequency. These different models were chosen because the relation between latency and MUAP size is unknown in diseased nerves. Model 1 is likely to reflect the relation between latency and MUAP size in healthy subjects (Kadrie and Brown, 1978; Dengler et al., 1988). Model 4 may reflect this relation in nerves in which conduction slowing due to demyelination or axonal degeneration occurs randomly among motor axons. After shifting MUAPs in time according to their assigned latency, all MUAPs within a set were arithmetically summated to obtain a distal CMAP. This procedure was repeated for each TD from 0 to 20 ms, for each of the four models of latency assignment and for each of the 12 MUAP sets. For model 1, each simulation was carried out once for each TD, yielding 21 CMAPs per MUAP set. For model 2, 3 and 4, latency assignment occurred randomly and, therefore, each simulation was repeated 10 times, thus yielding 210 CMAPs for each model and for each MUAP set.
To simulate a proximal CMAP, each MUAP that contributed to a distal CMAP was assigned an additional latency in the forearm segment according to the methods described for the distal segment: the latency of the fastest MUAP was fixed at 5 ms corresponding to a conduction velocity of 60 m/s over 30 cm, the latency of the slowest MUAP was 0–20 ms slower than the latency of the fastest MUAP, and the latencies of the remaining MUAPs were assigned according to the four models. Latency assignment within the forearm segment was independent of latency assignment in the distal segment. For each MUAP, the assigned latencies in the forearm and distal segments were added. After shifting each MUAP in time according to this cumulated latency, all MUAPs within a set were arithmetically summated to obtain a proximal CMAP. For model 1, one proximal CMAP was simulated for each distal CMAP and for each TD from 0 to 20 ms in the forearm segment, thus yielding 21 proximal CMAPs for each distal CMAP. For the other 3 models, each simulation was repeated 10 times (to account for the random latency assignment in these models), thus, yielding 210 proximal CMAPs for each distal CMAP. Altogether, 44 100 proximal CMAPs were simulated for each of the models 2, 3 and 4 and for each MUAP set. The model yielding the greatest decrement will be used to define criteria for CB. This ensures that the proposed criteria also hold true for other latency distributions since these will yield less decrement.
The above described models do not take into account that conduction within individual axons of a nerve may not be uniformly slowed. This is because, for a given axon, only the total latency of a MUAP between two stimulation sites of a segment (e.g. wrist-elbow or a shorter segment) is of interest. Whether this latency is related to uniform or multifocal slowing in the axons within the segment is irrelevant. The same principle holds true for slowing in the distal segment and its influence on decrement.
CMAP variables
For each simulated CMAP, we determined the amplitude, area and duration of the first negative phase, the total negative phase and the total CMAP. These variables are defined in Fig. 1. To determine the CMAP change in the forearm segment, each simulated proximal CMAP was compared with one simulated distal CMAP. The distal and proximal CMAPs were reconstructed from the same set of MUAPs and had identical TDs in the distal segment.
|
The following CMAP variables were determined for the first negative phase, for the total negative phase and for the total CMAP: amplitude decrement (%) calculated as (distal CMAP amplitude – proximal CMAP amplitude) x 100/(distal CMAP amplitude); area decrement (%) calculated as (distal CMAP area – proximal CMAP area) x 100/(distal CMAP area); duration prolongation (ms), calculated as (proximal CMAP duration – distal CMAP duration) and distal duration (ms) ( = duration of the distal CMAP). A biphasic CMAP was defined as a CMAP with one baseline crossing, and a polyphasic CMAP was defined as a CMAP with more than 1 baseline crossing.
Simulation of CB
The effect of CB on decrement was studied for the forearm segment. Simulations of distal and proximal CMAPs were similar to those described for forearm segments. For each proximal CMAP, we randomly blocked 0, 10, 20, 30, 40, 50, 60, 70, 80, 90 or 100% of axons. For each distal CMAP, for each set of MUAPs, for each percentage of blocked axons and for each of the 21 TDs between 0 and 20 ms, a proximal CMAP was simulated.
Assessment of the maximal TD in patients with CIDP
The difference between the latency of the slowest and that of the fastest axon was determined in the forearm segment of the median nerve using Ingram's single and double collision techniques (Ingram et al., 1987a, b). By combining the results of the single and double collision techniques, the latencies of the fastest and the slowest conducting axon within the median nerve were estimated assuming that fast conducting axons have shorter refractory periods than slow conducting axons (Rutten et al., 1998). For each patient, TD was calculated by subtracting the minimal latency from the maximal latency. To correct for the length of the nerve segments, each TD was divided by the length of the nerve segment (cm) in which it was determined and then multiplied by 30 so that standardized values for nerve segments with a length of 30 cm were obtained. The greatest TD found among median nerves of the five patients with CIDP was considered the maximal TD for the forearm segment of the median nerve that may occur as a result of immune-mediated demyelination.
Validation of proposed criteria
The proposed criteria were retrospectively validated in 20 patients with MMN and in 20 patients with LMND. Patients with MMN had an asymmetric lower motor neurone syndrome without sensory abnormalities, motor CB in at least one nerve segment on extensive nerve conduction studies (Van Asseldonk et al., 2003), normal sensory conduction and a favourable response to high-dose intravenous immunoglobulins (Van den Berg-Vos et al., 2000). CB was defined as an area decrement of the first negative phase of at least 50% (definite CB; Rhee et al., 1990) or an amplitude decrement of the first negative phase of at least 30% in an arm nerve (possible CB; Albers et al., 1985; Van Asseldonk et al., 2003). Patients with LMND had a slowly progressive lower motor neurone syndrome with adult onset, evidence of lower motor neurone involvement on neurological examination, electrophysiological evidence of lower motor neurone involvement on needle EMG (Van den Berg-Vos et al., 2003), no abnormalities of sensory conduction, no motor CB on extensive nerve conduction studies (Van Asseldonk et al., 2003), no clinical signs of upper motor neurone involvement and no structural lesions that could account for the clinical findings on MRI or myelography of the spinal cord (Van den Berg-Vos et al., 2003).
The diagnostic yield for CB of the proposed criteria was compared with that of the criterion proposed by Rhee et al. (1990) (area decrement of the first negative CMAP phase of at least 50%). The criteria were applied bilaterally to the forearm, upper arm and shoulder segments of the median nerve to the thenar and the ulnar nerve to the hypothenar. The finding of CB was scored as true positive in MMN. The absence of CB was scored as true negative in LMND. We calculated the sensitivity (number of true positive patients/number of patients with MMN) and the specificity (number of true negative patients/number of patients with LMND).
|
Results |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Characteristics of surface MUAPS
The characteristics of the recorded MUAPs are shown in Table 1. The number of recorded MUAPs ranged from 3 to 15 per muscle. Within and between MUAP sets, a wide variety of values were found especially for the amplitude. In general, the duration of the negative phase was considerably shorter than that of the positive phase. The percentage of polyphasic needle-recorded MUAPs varied between 7 and 50% among MUAP sets. Polyphasic MUAPs were not found on surface recording as each surface MUAP had one negative and one positive phase. Also at higher levels of effort, averaging resulted in clearly definable MUAPs.
|
CMAP reconstruction
To reconstruct the CMAP recorded after stimulation at the wrist, the number of surface MUAPs recorded from a muscle had to be multiplied 1–14 times until the amplitude of the negative phase of the simulated CMAP was similar to that of the recorded CMAP (Table 1). The number of MUAPs that were used for this reconstruction ranged from 3 to 124 per set. For each MUAP set, the amplitude and duration of the simulated CMAP were compared with those of the recorded CMAP (Table 2). The duration of the negative phase of the recorded CMAP was roughly similar to that of the simulated CMAP. The amplitude of the positive phase of the recorded CMAP was similar or smaller as compared with the simulated CMAP. The duration of the positive phase of the recorded CMAP was similar or greater as compared with the simulated CMAP.
|
Relation between latency distribution and decrement
To determine which of the four models of latency assignment yielded the greatest decrement in the forearm segment, the relation between decrement and TD in the forearm segment was studied without TD in the distal segment (Fig. 2). For each of the four models of latency assignment, the area decrement of the total negative phase increased when TD increased. This increase was most pronounced for model 4. Similar findings were obtained for the area and amplitude of the total negative phase and the total CMAP. The findings for the first negative phase are discussed below. Because model 4 resulted in the greatest decrement for most variables, this model was chosen for all further simulations.
|
Effect of TD in the forearm segment on CMAP decrement in the forearm segment
Figure 3 shows the relations between decrement, TD and duration prolongation in the forearm segment. For this analysis, TD in the distal segment was set at 0 ms. These relations will only be described for the area of the total negative phase and for the area of the first negative phase; the relations described for area were similar to those of amplitude and the relations described for the total negative phase were similar to those of the total CMAP. Each dot in Fig. 3 represents one simulation: black dots represent the CMAP change when the proximal CMAP was biphasic; grey dots represent the CMAP change when the proximal CMAP was polyphasic.
|
When measured from the total negative phase, relations were similar whether the proximal CMAP was biphasic or polyphasic. The decrement was found to increase with TD (Fig. 3A). The upper border formed by the dots in Fig. 3A represents the maximal decrement that may arise from different values of TD. However, it is not useful to derive criteria for CB from this relation because nerve conduction studies do not give direct information on TD. For this reason we determined whether TD is reflected by duration prolongation, a variable which can be readily measured from nerve conduction studies. Duration prolongation of the total negative phase increased linearly with TD (Fig. 3B). This indicates that TD is reflected by duration prolongation of the total negative phase. The upper border formed by the dots in Fig. 3C represents the maximal decrement of the total negative phase for different values of duration prolongation. The maximal decrement increased with duration prolongation, the increase becoming less prominent with higher values of duration prolongation and reaching a value of
70% at a duration prolongation between 15 and 20 ms (Fig. 3C). When measured from the first negative phase, the above-described relations were also present when the proximal CMAPs were biphasic. However, when the proximal CMAPs were polyphasic, the values were widely scattered; even for relatively small values of duration prolongation, the decrement could be considerable (Fig. 3D, E and F). For this reason the first negative phase was not used for further analysis.
Occurrence of polyphasic CMAPS
It was determined whether the occurrence of polyphasic proximal CMAPs depends on TD in the forearm segment, on the number of MUAPs that contribute to a CMAP or on both. For this analysis, TD in the distal segment was set at 0 ms. The percentage of polyphasic proximal CMAPs was plotted against TD in the forearm segment for CMAPs consisting of fewer than 30 MUAPs and for CMAPs consisting of more than 30 MUAPs (Fig. 4). As TD increased, the percentage of polyphasic proximal CMAPs increased. This was more pronounced for CMAPs that consisted of <30 MUAPs (MUAP sets 1–7) than for CMAPs that consisted of >30 MUAPs (MUAP sets 8–12), indicating that the occurrence of polyphasic proximal CMAPs increases owing to TD and to a low number of MUAPs contributing to the CMAP.
|
Effect of TD in the distal segment on decrement in the forearm segment
We studied whether TD in the distal segment was reflected by distal duration, because nerve conduction studies do not give information on distal TD. For all analyses in which distal duration was used, we used the duration of the total negative phase, because it allows more precise demarcation in nerve conduction studies than the duration of the total CMAP. Figure 5 shows the relation between distal duration and TD in the distal segment. Distal duration increased linearly with TD indicating that distal duration of the total negative CMAP reflects TD in the distal segment.
|
We assessed whether the maximal decrement that may arise from TD in the forearm segment is influenced by TD in the distal segment. Figure 6 shows the decrement in the forearm segment plotted against distal duration. These relations will only be described for the area of the total negative phase, as they were similar to those of total negative phase amplitude, total CMAP area and total CMAP amplitude. The upper border formed by the dots in Fig. 6 represents the maximal decrement for different values of distal duration. The greatest decrement (69%) was found when the distal duration was smallest (4 ms). If the distal duration increased from 4 to 12 ms, the maximal decrement decreased from 69 to
45%. For a distal duration between 12 and 20 ms, the maximal decrement was between 40 and 50%. These findings indicate that the maximal decrement that may arise from TD within the forearm segment decreases when TD in the distal segment increases.
|
Sensitivity for TD and for CB of different variables for decrement
To determine which variable for decrement was least sensitive for TD, we studied the relation between maximal decrement and duration prolongation in the forearm segment (Fig. 7). This was done for different values of distal duration. Each dot in Fig. 7 represents 99% of the maximum value of the decrement that was found for a specific combination of duration prolongation and distal duration values. This relation is only shown for the area and amplitude decrement of the total negative phase as a similar relation was found for the total CMAP. The maximal decrement increased as duration prolongation increased. This increase in maximal decrement with duration prolongation as well as the maximal decrement itself was most pronounced if distal duration was 9 ms or less, less pronounced if it was between 9 and 12 ms and least pronounced if it was 12 ms or more. The maximal decrement was less pronounced for area than for amplitude indicating that area decrement is less sensitive for TD than amplitude decrement.
|
To assess which variable for decrement was most sensitive for the detection of CB, we determined, for each CMAP variable and for each percentage of blocked axons between 0 and 100%, whether CB was detected. CB was assumed to be detected when decrement exceeded the 99th percentile of the maximal decrement that could occur owing to increased TD. Whether CB was detected was determined for the different variables and for each combination of duration prolongation and distal duration shown in Fig. 7. In this analysis, the distal duration of the total negative phase was used. For each CMAP variable studied, CB was detected more frequently as the percentage of blocked axons increased (Fig. 8). For each percentage of blocked axons between 10 and 80%, CB was detected more frequently for area decrement than for amplitude decrement. CB was detected with similar frequency for area decrement of the total negative phase and area decrement of the total CMAP (Fig. 8). These findings indicate that, whether measured from the total negative phase or from the total CMAP, area decrement is more sensitive for detection of CB than amplitude decrement.
|
Proposed criteria for CB
The proposed criteria for CB were derived from the 99th percentile of the maximal area decrement as shown in Fig. 7. The criteria are based on the assumption that it is likely that CB occurred in at least some axons in a forearm segment of the median nerve, if the decrement found on a nerve conduction study exceeds the decrement shown in Fig. 7, for a given combination of distal duration and duration prolongation. These criteria are more liberal if TD in the forearm segment is limited or TD in the distal segment is pronounced. Table 3 shows the proposed criteria for different combinations of distal duration and duration prolongation. The proposed criteria were based on area decrement because this was less influenced by TD and more by CB than amplitude decrement. Criteria were defined for decrement of the total negative phase and were rounded off to a multiple of 5. These criteria are also valid for decrement of the total CMAP because the 99th percentile decrement of the total CMAP differed by <2% from that of the total negative phase for each combination of duration prolongation and distal duration.
|
Maximal TD in patients with CIDP
The collision studies showed that the greatest TD obtained in CIDP patients was 16.5 ms for a conduction distance of 30 cm. This indicates that the maximal TD due to demyelination in CIDP was within the range of simulations of the present study. Simulation of TD in a forearm segment with a conduction distance of 30 cm showed that a TD of 16.5 ms corresponds to a duration prolongation of 16.5 ms or less (Fig. 3B). At this duration prolongation, the 99th percentile of maximal simulated area decrement was 69% for the total negative phase (Fig. 7) and 70% for the total CMAP (data not shown).
Validation of proposed criteria
Table 4 shows the diagnostic yield for CB of the proposed criteria and that of the criterion of Rhee et al. in patients with MMN and in patients with LMND. When bilaterally applied to the forearm segment of the median nerve, our criteria revealed eight segments with CB in six patients with MMN and the criterion of Rhee et al. revealed five segments with CB in four patients with MMN. When bilaterally applied to the forearm, upper arm and shoulder segments of the median nerve and the ulnar nerve, our criteria revealed 31 segments with CB in 17 patients with MMN and the criterion of Rhee et al. revealed 19 segments with CB in 13 patients with MMN. In patients with LMND, our criteria did not reveal CB and also the criterion of Rhee et al. did not reveal CB. In three patients without CB in the median or ulnar nerve the diagnosis of MMN had been made on the basis of nerve conduction studies in arm nerves innervating proximal muscles. As compared with the criterion of Rhee et al. our criteria increased the sensitivity for CB in the median or ulnar nerve of patients with MMN from 65 to 85%, whereas specificity in LMND remained 100%.
|
|
Discussion |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
The criteria for CB proposed in the present study were based on computer simulations of nerve conduction with data recorded in humans. These criteria are based on the assumption that, when a decrement, as found in a nerve conduction study, is larger than can be explained by increased TD or collateral reinnervation, there will be CB in at least some axons (Rhee et al., 1990
). The effects of these factors on decrement were studied by simulation of conduction in the median nerve using surface MUAPs recorded from the m. abductor pollicis brevis in patients with LMND. Surface MUAPs were used because the CMAP is the summation of surface MUAPs. MUAPs were recorded in patients with LMND to take the effect of MUAP shape due to collateral reinnervation on decrement into account. The effect of TD on decrement was calculated for a wide range of TD in the forearm segment and the segment distal to the forearm. The proposed criteria for CB were based on area decrement because this was less influenced by TD and more by CB than amplitude decrement. The maximal area decrement in the forearm that could be attained due to TD in the forearm increased as TD in the forearm segment increased but decreased as TD in the distal segment increased. We showed that duration prolongation reflected TD within the forearm segment and that distal duration reflected TD in the distal segment. As, contrary to TD, duration prolongation and distal duration are readily available from nerve conduction studies, they were used to assess TD. Thus, less rigid criteria could be proposed for nerve segments with limited duration prolongation in the forearm or pronounced distal duration. A separate investigation showed that the maximal amount of TD in patients with CIDP was well within the range of our simulations, indicating that the range of simulations was realistic. The proposed criteria were validated by applying them retrospectively to patients with MMN and patients with LMND in whom MMN was excluded. The proposed criteria were more sensitive and equally specific for CB as compared with the criterion for CB based on the simulation study using animal data (Rhee et al., 1990). This was true for median nerve forearm segments (from which our criteria were derived) and also for median nerve upper arm and shoulder segments and for ulnar nerve segments.
The fact that our criteria are based on MUAPs affected by reinnervation raises the question about whether our criteria can be applied in disorders in which nerves are not affected by reinnervation such as acute Guillain Barré syndrome. As compared with surface MUAPs recorded from the m. abductor pollicis brevis in healthy subjects, some of our surface-recorded MUAPs had a similar size whereas others had a greater size (Slawnych et al., 1990; Stein and Yang, 1990; McComas, 1991). As a consequence, the range of MUAP sizes used in the present study was, for some MUAP sets, greater than those in healthy subjects and, most likely, also greater than those of patients with acute Guillain Barré syndrome. Whether the MUAP size range is a determinant of decrement was not investigated in our study, but it was investigated in a previous simulation study by Rhee et al. (1990). Using data in healthy rats, Rhee et al. showed that the maximal area decrement of CMAPs consisting of MUAPs with similar size was not greater than that of CMAPs consisting of a large range of MUAP sizes. This may suggest that our criteria are also applicable in nerves in which motor units are unaltered.
Unexpectedly, the surface MUAPs that were used for CMAP reconstruction were not polyphasic despite the fact that the associated needle MUAPs could be polyphasic owing to collateral reinnervation. This indicates that the suggested mechanism of increased phase cancellation of polyphasic MUAPs contributing to the CMAP does not contribute to decrement since the CMAP is built up from non-polyphasic surface MUAPs. However, other mechanisms related to collateral reinnervation, not previously reported, were found to affect decrement. Some of the surface-recorded MUAPs in the present study had a much greater size than surface MUAPs recorded from the m. abductor pollicis brevis in healthy subjects (Slawnych et al., 1990; Stein and Yang, 1990; McComas, 1991). Large surface MUAPs most probably reflect partial denervation followed by collateral reinnervation (Doherty et al., 1995; Wang and Delwaide, 1998). Large surface MUAPs limit the number of MUAPs required for CMAP reconstruction (Slawnych et al., 1990; Doherty et al., 1995). The present study showed that CMAPs consisting of a low number of MUAPs are more likely to become polyphasic when TD increases as compared with CMAPs consisting of a high number of MUAPs. It is likely that polyphasic CMAPs were not found in previous simulation studies because CMAPs were reconstructed from a high number of small MUAPs recorded in healthy rats or humans (Lee et al., 1975; Rhee et al., 1990). In the present study, simulations with polyphasic CMAPs showed that TD cannot be estimated from the first negative phase when the proximal CMAP is polyphasic. This precludes the use of the decrement of this variable for the assessment of CB. Because polyphasic CMAPs were reported in patients with demyelinating polyneuropathies as well as in patients with motor neurone disease, criteria for CB are only useful when polyphasic CMAPs can be taken into account (Thaisetthawatkul et al., 2002). To allow application to biphasic as well as polyphasic CMAPs, the proposed criteria were based on decrement of the total negative phase or total CMAP.
The maximal conduction velocity and the range of conduction velocities may influence TD and, therefore, the maximal decrement due to TD. When the maximal conduction velocity remains the same and the range of conduction velocities increases, TD will increase; when the maximal conduction velocity decreases and the range of conduction velocities remains the same, TD will also increase. For conduction velocities of 60, 50, 40 and 30 m/s, the latencies over 7 cm are 1.17, 1.40, 1.75 and 2.33 ms, respectively. Thus, TD is smaller for a conduction velocity range of 60–40 m/s (0.58 ms) than for a conduction velocity range of 50–30 m/s (0.93 ms). Therefore, our relatively high maximal distal conduction velocity of 60 m/s for the fastest MUAP could have introduced errors due to a relatively small TD. However, since we simulated ranges of TD leading to maximal decrement in the distal as well as in the forearm segment, our results apply to any degree of TD and to any combination of conduction velocities.
As compared with the reconstructed CMAP, the recorded CMAPs had a smaller positive phase amplitude and a greater duration. This was also found for the CMAPs reconstructed with surface MUAPs recorded from the m. abductor pollicis brevis in healthy subjects (Lee et al., 1975). It was suggested that the positive phase of the CMAP recorded from the m. abductor pollicis brevis was reduced owing to phase cancellation by a contribution from the first two lumbrical muscles, which are also innervated by the median nerve. In CMAP recordings from the hypothenar, muscles not directly under the active electrode site may contribute a major component to the CMAP (McGill and Lateva, 1999). The size of this contribution depends on temperature as well as finger position (McGill and Lateva, 1999). To keep this contribution constant, MUAP and CMAP recordings in the present study were performed at a standardized temperature of 37°C (Franssen and Wieneke, 1994) and fingers were fixated. Moreover, the difference between the size of the recorded and that of the reconstructed CMAP in the present study can be explained by the fact that the recorded CMAP is affected by TD over 7 cm, whereas the reconstructed CMAP at 0 cm is not affected by TD. This difference will decrease the amplitude and increase the duration of the recorded as compared with the reconstructed CMAP. The reconstruction of the CMAP with a distal TD of zero may have led to an erroneously small number of MUAPs needed for this reconstruction, since phase cancellation is minimal with a TD of zero. However, in the patients with LMND from whom the MUAPs were obtainted, TD over 7 cm was probably small as these patients had no demyelination. We therefore compared the number of MUAPs needed to reconstruct CMAPs with TDs of 0.00 ms, 0.23 ms which we found previously in the median nerve of normal subjects (Rutten et al., 1998), 0.58 and 0.93 ms (see above). We found that the median additional numbers of MUAPs needed to reconstruct the CMAP were 1, 2 and 3 for distal TDs of 0.23, 0.58 and 0.93 ms, respectively. These additional numbers of MUAPs are small as compared with the range of numbers of MUAPs needed to reconstruct the CMAP with a distal TD of zero (3–124, see Table 1). Such a small increase is unlikely to have affected the results of our simulations. On the other hand, the findings in the present study indicated that CMAP reconstruction with distal TD would have resulted in criteria that were not specific enough for CB as compared with CMAP reconstruction without distal TD. This is because distal TD decreases the maximal decrement that may occur in the forearm due to TD in the forearm.
Our simulations showed that area decrement of the total negative phase as well as area decrement of the total CMAP were equally sensitive for detection of CB. However, not all available recording apparatus allows for the demarcation of the area of the total negative phase. Thus, the choice of which CMAP variables should be used for detection of CB depends on the CMAP registration software and on the preference of the electromyographer. To allow proposal of a more flexible set of criteria, our criteria for CB were specified for different values of distal duration and duration prolongation. For this purpose, we used the duration of the total negative phase, which can be demarcated more precisely than the duration of the total CMAP and which can be demarcated on all recording apparatus.
For the development of our criteria, we used the maximum distance for the forearm segment (30 cm) that was measured in our population. This was not because our criteria were meant to be applied only to forearm segments of 30 cm but to simulate the maximum possible effect of distance in a forearm segment on TD. However, TD not only increases with increasing conduction distance but also with increasing demyelination. To account for the latter, we simulated increasing TDs over a conduction distance of 30 cm. This enabled us to simulate the maximum possible TD (and decrement due to TD) over a distance of 30 cm. By comparing the simulated TD over 30 cm with the measured TD corrected for 30 cm in patients with CIDP, we could show that the simulated TDs were realistic. Nevertheless, whether TD arises from a large conduction distance or different conduction velocities due to demyelination or any combination of these was irrelevant for the development of our criteria for CB. This is because our criteria depend on the amount of TD only and not on whether this TD is due to distance or demyelination. Therefore, our criteria are also applicable to conduction distances of <30 cm.
The finding that the maximal decrement in the forearm segment due to TD in the forearm segment decreased when TD in the distal segment increased can be explained by the following mechanism. If there is no distal TD, TD in the forearm leads to decrement through desynchronization and phase cancellation. On the other hand, with considerable distal TD (leading to distal desynchronization and distal phase cancellation), TD in the forearm segment leads to decrement mainly through desynchronization and less through phase cancellation. This is because the phase cancellation that has already occurred in the distal segment cannot occur anymore in the forearm segment; therefore, less decrement can occur due to TD in the forearm segment. This mechanism has implications for application of the proposed criteria to upper arm or shoulder segments. The distal segment is measured from the distal site of stimulation to the muscle from which the CMAP is recorded. Thus, when conduction is measured in the upper arm segment of the median nerve, the distal segment is the segment from the elbow to the m. abductor pollicis brevis. Consequently, TD within this long segment decreases the maximal decrement that may occur due to TD in the upper arm. This implies that more liberal criteria are applicable to the upper arm when TD distal to the elbow is increased. The finding that distal TD was reflected by distal duration allowed proposal of criteria for CB that take distal TD into account, irrespective of the length of the distal segment. For instance, when studying the upper arm segment, distal TD is assessed from the duration of the CMAP on stimulation of the elbow. Our criteria may, therefore, also be applicable to the forearm, upper arm and shoulder segments of the median nerve.
It should be emphasized that the criteria for CB derived in the present study using MUAPs recorded from the m. abductor pollicis brevis cannot simply be applied to other nerves. The number of motor units within a muscle and the size of the MUAPs that contribute to the CMAP affect the maximal decrement that may occur due to TD (Rhee et al., 1990; McComas, 1991; Brown and Bolton, 1993); these variables differ among muscles (McComas, 1991; Rhee et al., 1990; Brown and Bolton, 1993). In healthy subjects, the number of motor units and the size of MUAPs of the m. abductor pollicis brevis are different for muscles in the upper arm or leg but are similar to those of the m. abductor digiti V (Slawnych et al., 1990; McComas, 1991). It is, therefore, conceivable that our criteria for CB are also applicable to CMAPs recorded from the m. abductor digiti V although no data were obtained for the m. abductor digiti V in the present study.
At present, there is no reference test by means of which CB as detected by nerve conduction studies can be validated. The only way to determine whether the proposed criteria are real is to determine whether they reveal CB in disorders in which CB is likely, such as MMN, and do not reveal CB in disorders in which CB is unlikely, such as LMND. Our retrospective validation study showed that our proposed criteria for CB improved sensitivity for CB in MMN as compared with the criterion for CB based on the simulation study using animal data (Rhee et al., 1990). This may have been due to the use of the total negative phase, and not the first negative phase, to detect CB and due to the fact that our criteria are more liberal when TD is limited. It should be emphasized that our criteria were derived from data of the forearm segment of the median nerve and that they are, therefore strictly speaking, only applicable to that segment. However, the arguments presented in the previous two paragraphs favour their application to all segments of the median and ulnar nerve to thenar and hypothenar muscles. Moreover, the finding that our criteria revealed CB in forearm, upper arm and shoulder segments of the median and the ulnar nerves of patients with MMN, but not in those of patients with LMND, suggests that they are specific for CB.
Future prospective validation studies are required to determine whether our criteria may improve accuracy of diagnosis of patients with potentially treatable immune-mediated polyneuropathies as compared with criteria for CB that were based on methods other than simulation studies (e.g. Cappellari et al., 1997; Ghosh et al., 2005). In these future studies, consecutive patients suspected of MMN should be included and a positive response to intravenous immunoglobulins used as a reference standard.
|
Acknowledgements |
|---|
We thank Tineke Gebbink for her excellent technical support with MUAP recordings and collision techniques. Supported by a grant from the Prinses Beatrix Fonds.
|
References |
|---|
|
TopSummaryIntroductionPatients and methodsResultsDiscussionReferences |
|---|
Albers JW and Kelly JJ Jr. (1989) Acquired inflammatory demyelinating polyneuropathies: clinical and electrodiagnostic features. Muscle Nerve 12:435–51.[CrossRef][Web of Science][Medline]
Albers JW, Donofrio PD, Mc Gonagle TK. (1985) Sequential electrodiagnostic abnormalities in acute inflammatory demyelinating polyradiculoneuropathie. Muscle Nerve 8:528–39.[CrossRef][Web of Science][Medline]
American Association of Electrodiagnostic Medicine. (1999) Consensus criteria for the diagnosis of partial conduction block. Muscle Nerve 22:s225–9.
American Association of Neurology. (1991) Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology 41:617–8.
Berger AR, Bradley WG, Brannagan TH, Busis NA, Cros DP, Dalakas MC, et al. (2003) Guidelines for the diagnosis and treatment of chronic inflammatory demyelinating polyneuropathy. J Peripher Nerv Syst 8:282–4.[Web of Science][Medline]
Biro G and Partridge LD. (1971) Analysis of multiunit spike records. J Appl Physiol 30:521–6.
In Brown WF and Bolton CF (Eds.). Clinical electromyography (1993) (Butterworth-Heinemann, Boston).
Brown WF and Feasby TE. (1984) Conduction block and denervation in Guillain-Barre polyneuropathy. Brain 107:219–39.
Cappellari A, Nobile-Orazio E, Meucci N, Levi MG, Scarlato G, Barbieri S. (1997) Criteria for early detection of conduction block in multifocal motor neuropathy (MMN): a study based on control populations and follow-up of MMN patients. J Neurol 244:625–30.[CrossRef][Web of Science][Medline]
Chaudhry V, Corse AM, Cornblath DR, Kuncl RW, Drachman DB, Freimer ML, et al. (1993) Multifocal motor neuropathy: response to human immune globulin. Ann Neurol 33:237–42.[CrossRef][Web of Science][Medline]
Cornblath DR. (1990) Electrophysiology in Guillain-Barre syndrome. Ann Neurol 27:Suppl, S17–20.
Cornblath DR, Sumner AJ, Daube J, Gilliat RW, Brown WF, Parry GJ, et al. (1991) Conduction block in clinical practice. Muscle Nerve 14:869–71.[CrossRef][Web of Science][Medline]
Dalakas MC. (2002) Mechanisms of action of IVIg and therapeutic considerations in the treatment of acute and chronic demyelinating neuropathies. Neurology 59:S13–21.
Dengler R, Stein RB, Thomas CK. (1988) Axonal conduction velocity and force of single human motor units. Muscle Nerve 11:136–45.[CrossRef][Web of Science][Medline]
Doherty T, Simmons Z, O'Connell B, Felice KJ, Conwit R, Chan KM, et al. (1995) Methods for estimating the numbers of motor units in human muscles. J Clin Neurophysiol 12:565–84.[Web of Science][Medline]
Dyck PJ, O'Brien PC, Oviatt KF, Dinapoli RP, Daube JR, Bartleson JD, et al. (1982) Prednisone improves chronic inflammatory demyelinating polyradiculoneuropathy more than no treatment. Ann Neurol 11:136–41.[CrossRef][Web of Science][Medline]
Franssen H and Wieneke GH. (1994) Nerve conduction and temperature: necessary warming time. Muscle Nerve 17:336–44.[CrossRef][Web of Science][Medline]
Ghosh A, Busby M, Kennett R, Mills K, Donaghy M. (2005) A practical definition of conduction block in IvIg responsive multifocal motor neuropathy. J Neurol Neurosurg Psychiatry 76:1264–8.
Hahn AF, Bolton CF, Pillay N, Chalk C, Benstead T, Bril V, et al. (1996a) Plasma-exchange therapy in chronic inflammatory demyelinating polyneuropathy. A double-blind, sham-controlled, cross-over study. Brain 119:1055–66.
Hahn AF, Bolton CF, Zochodne D, Feasby TE. (1996b) Intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy. A double-blind, placebo-controlled, cross-over study. Brain 119:1067–77.
Hughes R, Bensa S, Willison H, Van den BP, Comi G, Illa I, et al. (2001) Randomized controlled trial of intravenous immunoglobulin versus oral prednisolone in chronic inflammatory demyelinating polyradiculoneuropathy. Ann Neurol 50:195–201.[CrossRef][Web of Science][Medline]
Ingram DA, Davis GR, Swash M. (1987a) The double collision technique: a new method for measurement of the motor nerve refractory period distribution in man. Electroencephalogr Clin Neurophysiol 66:225–34.[CrossRef][Web of Science][Medline]
Ingram DA, Davis GR, Swash M. (1987b) Motor nerve conduction velocity distributions in man: results of a new computer-based collision technique. Electroencephalogr Clin Neurophysiol 66:235–43.[CrossRef][Web of Science][Medline]
Kadrie HA and Brown WF. (1978) Neuromuscular transmission in human single motor units. J Neurol Neurosurg Psychiatry 41:193–204.
Kaji R. (2003) Physiology of conduction block in multifocal motor neuropathy and other demyelinating neuropathies. Muscle Nerve 27:285–96.[CrossRef][Web of Science][Medline]
Kaji R, Shibasaki H, Kimura J. (1992) Multifocal demyelinating motor neuropathy: cranial nerve involvement and immunoglobulin therapy. Neurology 42:506–9.
Lee RG, Ashby P, White DG, Aguayo AJ. (1975) Analysis of motor conduction velocity in the human median nerve by computer simulation of compound muscle action potentials. Electroencephalogr Clin Neurophysiol 39:225–37.[CrossRef][Web of Science][Medline]
Lewis RA, Sumner AJ, Brown MJ, Asbury AK. (1982) Multifocal demyelinating neuropathy with persistent conduction block. Neurology 32:958–64.
McComas AJ. (1991) Invited review: motor unit estimation: methods, results, and present status. Muscle Nerve 14:585–97.[CrossRef][Web of Science][Medline]
McGill KC and Lateva ZC. (1999) The contribution of the interosseous muscles to the hypothenar compound muscle action potential. Muscle Nerve 22:6–15.[CrossRef][Web of Science][Medline]
Milner-Brown HS, Stein RB, Yemm R. (1973) The contractile properties of human motor units during voluntary isometric contractions. J Physiol 228:285–306.
Nandedkar S and Stalberg E. (1983) Simulation of macro EMG motor unit potentials. Electroencephalogr Clin Neurophysiol 56:52–62.[CrossRef][Web of Science][Medline]
Nobile-Orazio E, Meucci N, Barbieri S, Carpo M, Scarlato G. (1993) High-dose intravenous immunoglobulin therapy in multifocal motor neuropathy. Neurology 43:537–44.
Oh SJ, Kim DE, Kuruoglu HR. (1994) What is the best diagnostic index of conduction block and temporal dispersion? Muscle Nerve 17:189–93.
Olney RK, Budingen HJ, Miller RG. (1987) The effect of temporal dispersion on compound action potential area in human peripheral nerve. Muscle Nerve 10:728–33.[CrossRef][Web of Science][Medline]
Olney RK, Lewis RA, Putnam TD, Campellone JV Jr. (2003) Consensus criteria for the diagnosis of multifocal motor neuropathy. Muscle Nerve 27:117–21.[CrossRef][Web of Science][Medline]
Rhee EK, England JD, Sumner AJ. (1990) A computer simulation of conduction block: effects produced by actual block versus interphase cancellation. Ann Neurol 28:146–56.[CrossRef][Web of Science][Medline]
Rutten GJ, Gaasbeek RD, Franssen H. (1998) Decrease in nerve temperature: a model for increased temporal dispersion. Electroencephalogr Clin Neurophysiol 109:15–23.[CrossRef][Medline]
Slawnych MP, Laszlo CA, Hershler C. (1990) A review of techniques employed to estimate the number of motor units in a muscle. Muscle Nerve 13:1050–64.[CrossRef][Web of Science][Medline]
Stein RB and Yang JF. (1990) Methods for estimating the number of motor units in human muscles. Ann Neurol 28:487–95.[CrossRef][Web of Science][Medline]
Thaisetthawatkul P, Logigian EL, Herrmann DN. (2002) Dispersion of the distal compound muscle action potential as a diagnostic criterion for chronic inflammatory demyelinating polyneuropathy. Neurology 59:1526–32.
Van Asseldonk JTH, Van den Berg LH, Van den Berg-Vos RM, Wieneke GH, Wokke JH, Franssen H. (2003) Demyelination and axonal loss in multifocal motor neuropathy: distribution and relation to weakness. Brain 126:186–98.
Van den Berg-Vos RM, Franssen H, Wokke JH, Van Es HW, Van den Berg LH. (2000) Multifocal motor neuropathy: diagnostic criteria that predict the response to immunoglobulin treatment. Ann Neurol 48:919–26.[CrossRef][Web of Science][Medline]
Van den Berg-Vos RM, Franssen H, Wokke JH, Van den Berg LH. (2002) Multifocal motor neuropathy: long-term clinical and electrophysiological assessment of intravenous immunoglobulin maintenance treatment. Brain 125:1875–86.
Van den Berg-Vos RM, Visser J, Franssen H, de Visser M, de Jong JM, Kalmijn S, et al. (2003) Sporadic lower motor neuron disease with adult onset: classification of subtypes. Brain 126:1036–47.
Van Doorn PA, Brand A, Strengers PF, Meulstee J, Vermeulen M. (1990) High-dose intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy: a double-blind, placebo-controlled, crossover study. Neurology 40:209–12.
Wang FC and Delwaide PJ. (1998) Number and relative size of thenar motor units in ALS patients: application of the adapted multiple point stimulation method. Electroencephalogr Clin Neurophysiol 109:36–43.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||