**6.1. MSA**

Computational Intelligence in Electromyography Analysis – 294 A Perspective on Current Applications and Future Challenges

This figure illustrates where to insert concentric needles to measure external sphincter EMG.

A normal MUP usually has a 50–500 microV amplitude, a 3–8 msec duration, and 2–4 phases. In order to assess reinnervation, usually 10–20 single MUPs are recorded, which are automatically provided by an EMG computer. To ascertain single MUP, we still check each wave manually and adjust the onset and offset of each wave. It is particularly important to include late components (satellite potentials) to measure the duration of each unit.42 When the muscle is chronically denervated, an intact nerve tends to innervate the adjacent denervated muscle fibers. As a result, MUPs become of high amplitude, of long duration, and polyphasic. Among various EMG parameters, the use of duration, MUP area, and number of turns is recommended for optimal diagnostic power (sensitivity and specificity) in the EAS muscle.45 In addition, the results are dependent on the methods used; e.g., including or excluding late components. Palace et al. proposed that either of two criteria is sufficient to diagnose neurogenic changes in the EAS-EMG: (*a*) more than 20% of MUPs have a duration > 10 msec, or (*b*) the average duration of MUPs > 10 msec, particularly including the late components.38 When satellite potentials were excluded, the duration of MUPs did not differ significantly between Parkinson's disease and MSA.48 When lower motor neuron-type abnormalities are not apparent, it is reported that abnormal MUP recruitment pattern suggests pyramidal tract involvement.18 In addition to MUP analysis in the external sphincter muscles, other neurophysiologic tests, e.g., pudendal nerve conduction, sacral reflexes, somatosensory evoked potentials and cranial magnetic stimulation, and urodynamic studies, can be of particular value in the study of autonomic

**Figure 4.** The external anal sphincter and the external urethral sphincter.

patients.29,40,41

Cardiovascular autonomic failure in MSA is thought to derive from neuron loss in the thoracolumbar intermediolateral (IML) cell columns of the spinal cord and the medullary circulation center. In contrast, lower urinary tract disorder in MSA is thought to reflect multiple lesions in the basal ganglia and the pontine storage center (storage-facilitating areas), as well as in the pontine micturition center in or adjacent to the locus ceruleus and the sacral IML cell columns (voiding-facilitating areas).11 In addition, a distinguishing pathology in MSA is neuronal cell loss in the sacral Onuf nucleus.33,37

The first reports on neurogenic changes of EAS-EMG in MSA are attributed to Sakuta et al. (1978).62 Since then, EAS-EMG results for over 500 MSA patients have been reported, with abnormality rates of more than 70% in many studies5,30,36,38,47,53,62,64,71. EAS-EMG is better tolerated and yields identical results to those from EUS investigation5. Abnormalities have also been recorded in the bulbocavernosus muscles in MSA.67 In a larger study, Beck et al. (1994) reported that all (100%) 62 MSA patients with urological symptoms had abnormalities in both EAS and EUS-EMG.5 Palace et al. (1997) reported abnormal EAS-EMG in 103 (82%) of 126 patients with MSA38. Chandiramani et al. (1997) found abnormal EAS-EMG in 49 (94%) of 52 patients with MSA10. Kirchhof et al. (1999) found abnormal EAS-EMG in 89 (91%) of 98 patients with MSA28. Sakakibara et al. (2000) found an abnormal EAS-EMG in 53 (74%) of 71 MSA patients52. These abnormalities correspond to selective loss of ventral horn cells and astrogliosis; the loss is particularly severe in the second and third sacral segments (Onuf's nucleus) in MSA11. Sphincter EMG has been proposed as a means of distinguishing between MSA and idiopathic Parkinson's disease (as described below), since the anterior horn cells of Onuf's nucleus are not affected in idiopathic Parkinson's disease.10 In contrast, there have been debates about whether or not sphincter EMG can be used to distinguish MSA from idiopathic Parkinson's disease. In a study of 13 patients with idiopathic Parkinson's disease and 10 patients with MSA, Giladi et al. (2000) found significant overlap in all EMG parameters (presence of fibrillation potentials, MUP duration, presence of satellite potentials, percentage of polyphasic potentials)19. However, the durations of MUPs in both the MSA and Parkinson's disease groups were longer than in other studies.

It is reported that EAS-MUP abnormalities can distinguish MSA from idiopathic Parkinson's disease in the first 5 years after disease onset.30,69,74 However, the prevalence of such abnormalities in the early stages of MSA has not been well known. In our recent study of 84 probable MSA cases, 62% exhibited neurogenic change.80 The prevalence was relatively low presumably because up to 25% of our patients had a disease duration of 1 year or less. In such early cases, the diagnosis of MSA should be made with extreme caution. In addition to the clinical diagnostic criteria, we usually add an imaging study and we perform gene analysis to the extent possible. The prevalence of neurogenic change was 52% in the first year after disease onset, which increased to 83% by the fifth year (p<0.05) **(Fig. 5)**. Among the patients who underwent repeated studies, many had normal to mild abnormality at the initial examination, which turned into marked abnormality during the course of illness **(Fig. 6)**. Therefore, as expected, it is apparent that the involvement of Onuf's nucleus in MSA is time-dependent. In the early stages of illness, the prevalence of neurogenic change in MSA does not seem to be high. In 2 patients who underwent repeated studies, the EAS-EMG findings tended to remain normal. We do not know whether some MSA patients never develop neurogenic change during the course of their illness. However, Wenning et al. (1994) reported 3 patients with normal EAS-EMG and a postmortem confirmation of MSA.77 Therefore, a negative result cannot exclude a diagnosis of MSA. More recently, Paviour et al. (2005) reported that among 30 sets of clinical data and postmortem confirmation in MSA cases with a duration of more than 5 years, 24 (80%) had abnormal EAS-EMG, 5 (17%) had a borderline result, and only 1 had a normal EMG.39

Sphincter EMG for Diagnosing MSA and Related Disorders 297

Percentage of MUPs with duration > 10 msec: one of two categories for neurogenic sphincter EMG.

The prevalence of neurogenic change also increased with the severity of gait disturbance (p<0.05) in our study. However, neurogenic change was not related to postural hypotension (reflecting adrenergic nerve dysfunction); erectile dysfunction in men (presumably reflecting cholinergic and nitrate oxidergic nerve dysfunction); detrusor overactivity (reflecting the central type of detrusor dysfunction); constipation (presumably reflecting both peripheral and central types of autonomic and somatic dysfunction); or gender **(Table 1)**. The neurogenic change in EAS-MUP was slightly more common in those with detrusorsphincter dyssynergia (reflecting the central type of sphincter dysfunction). It has been reported that neurogenic change does not correlate directly with a clinically obvious functional deficit.74 Patients with marked abnormalities in EAS-MUP may have no faecal incontinence,77 although, in such patients, anal sphincter weakness is not uncommon.58 In our study, the prevalence of neurogenic change slightly increased with the severity of storage disorder (incontinence). The most common type of urinary incontinence in MSA is urgency incontinence, which results mostly from detrusor (bladder) overactivity. However, we noted urinary incontinence in 17 patients without detrusor overactivity or lowcompliance detrusor; in those cases, the urinary incontinence may have had a sphincter etiology. Urinary incontinence was more severe in the patients with neurogenic change than

We recently retrospectively analysed 445 case records of EMG cystometry with pressure flow studies, single motor unit potential (MUP) analysis in patients with parkinsonian syndrome, e.g., MSA: n=267, Parkinson's disease (PD): n=129, Dementia with Lewy bodies (DLB): n=25, and progressive supranuclear palsy (PSP): n=24. We carried out receiver operating characteristics (ROC) analysis, revealing that an area under the ROC curve (AUC) in differentiating MSA from other parkinsonian syndrome was 0.70 in duration, 0.62 in phase and 0.51 in amplitude, respectively, with statistically significance. Therefore, duration

of MUPs is most sensitive for the differentiation of MSA among MUPs parameters.

**Figure 6.** Percentage of MUPs with duration > 10 msec and duration of illness.

oblique figures: the number of patients who underwent the study repeatedly

straight figures: the number of patients

in those without it (p<0.05).

MUP: motor unit potential (cited from ref. 74)

The prevalence of neurogenic sphincter EMG increased during the course of illness. MUP: motor unit potential (cited from ref. 74)

**Figure 5.** Neurogenic sphincter EMG and duration of illness.

Percentage of MUPs with duration > 10 msec: one of two categories for neurogenic sphincter EMG. straight figures: the number of patients oblique figures: the number of patients who underwent the study repeatedly MUP: motor unit potential

(cited from ref. 74)

Computational Intelligence in Electromyography Analysis – 296 A Perspective on Current Applications and Future Challenges

borderline result, and only 1 had a normal EMG.39

The prevalence of neurogenic sphincter EMG increased during the course of illness.

**Figure 5.** Neurogenic sphincter EMG and duration of illness.

MUP: motor unit potential (cited from ref. 74)

year after disease onset, which increased to 83% by the fifth year (p<0.05) **(Fig. 5)**. Among the patients who underwent repeated studies, many had normal to mild abnormality at the initial examination, which turned into marked abnormality during the course of illness **(Fig. 6)**. Therefore, as expected, it is apparent that the involvement of Onuf's nucleus in MSA is time-dependent. In the early stages of illness, the prevalence of neurogenic change in MSA does not seem to be high. In 2 patients who underwent repeated studies, the EAS-EMG findings tended to remain normal. We do not know whether some MSA patients never develop neurogenic change during the course of their illness. However, Wenning et al. (1994) reported 3 patients with normal EAS-EMG and a postmortem confirmation of MSA.77 Therefore, a negative result cannot exclude a diagnosis of MSA. More recently, Paviour et al. (2005) reported that among 30 sets of clinical data and postmortem confirmation in MSA cases with a duration of more than 5 years, 24 (80%) had abnormal EAS-EMG, 5 (17%) had a

**Figure 6.** Percentage of MUPs with duration > 10 msec and duration of illness.

The prevalence of neurogenic change also increased with the severity of gait disturbance (p<0.05) in our study. However, neurogenic change was not related to postural hypotension (reflecting adrenergic nerve dysfunction); erectile dysfunction in men (presumably reflecting cholinergic and nitrate oxidergic nerve dysfunction); detrusor overactivity (reflecting the central type of detrusor dysfunction); constipation (presumably reflecting both peripheral and central types of autonomic and somatic dysfunction); or gender **(Table 1)**. The neurogenic change in EAS-MUP was slightly more common in those with detrusorsphincter dyssynergia (reflecting the central type of sphincter dysfunction). It has been reported that neurogenic change does not correlate directly with a clinically obvious functional deficit.74 Patients with marked abnormalities in EAS-MUP may have no faecal incontinence,77 although, in such patients, anal sphincter weakness is not uncommon.58 In our study, the prevalence of neurogenic change slightly increased with the severity of storage disorder (incontinence). The most common type of urinary incontinence in MSA is urgency incontinence, which results mostly from detrusor (bladder) overactivity. However, we noted urinary incontinence in 17 patients without detrusor overactivity or lowcompliance detrusor; in those cases, the urinary incontinence may have had a sphincter etiology. Urinary incontinence was more severe in the patients with neurogenic change than in those without it (p<0.05).

We recently retrospectively analysed 445 case records of EMG cystometry with pressure flow studies, single motor unit potential (MUP) analysis in patients with parkinsonian syndrome, e.g., MSA: n=267, Parkinson's disease (PD): n=129, Dementia with Lewy bodies (DLB): n=25, and progressive supranuclear palsy (PSP): n=24. We carried out receiver operating characteristics (ROC) analysis, revealing that an area under the ROC curve (AUC) in differentiating MSA from other parkinsonian syndrome was 0.70 in duration, 0.62 in phase and 0.51 in amplitude, respectively, with statistically significance. Therefore, duration of MUPs is most sensitive for the differentiation of MSA among MUPs parameters.


Sphincter EMG for Diagnosing MSA and Related Disorders 299

another feature. We performed urodynamic studies in 7 patients with DLB, and performed EAS-EMG in 3. Two of those 3 patients exhibited neurogenic changes in

AFPD is an intermediate entity that describes a combination of autonomic failure and IPD, but without dementia. We performed urodynamic studies in 7 patients with AFPD and performed EAS-EMG in 4. Three of those 4 patients exhibited neurogenic changes in

Earlier studies reported normal EAS-EMG in small groups of patients with PAF. However, Ravits et al. (1996)46 found abnormal EAS-EMG in 2 of 7 patients with PAF, although both of them were multiparous women. Sakakibara et al. performed urodynamic studies in 6 patients with PAF and EAS-EMG in 4. Three of those 4 patients exhibited neurogenic changes in MUPs.51 In PAF, parkinsonism may appear after a 10-year interval.81 Therefore PAF can be listed in the differential diagnosis of degenerative parkinsonism. To sum up, in all three Lewy body diseases (DLB, AFPD, PAF), the frequency of neurogenic changes seemed higher in EAS-EMG than in IPD but lower than in MSA. This suggests the involvement of the sacral Onuf's nucleus or its fibers in the external sphincter in these diseases. The prevalence of neurogenic changes in EAS-EMG seems to be: **MSA >> DLB = AFPD = PAF >> PD (Table 2)**. However, these assumptions require confirmation with a larger study. The results seem to be in accordance with the fact that 29% of the DLB patients undergoing EMG-cystometry had a low-compliance detrusor, indicating a pre-ganglionic lesion of the pelvic nerves. The bethanechol test showed that both of these patients had denervation supersensitivity of the detrusor, indicating a post-ganglionic lesion of the pelvic nerves. The results of physiological studies and metaiodobenzylguanidine (MIBG) cardiac scintigraphy suggested post-

MSA: multiple system atrophy, DLB: Dementia with Lewy bodies, AFPD: Autonomic failure with Parkinson's disease

**Table 2.** Comparison of lower urinary tract function in DLB, AFPD, PAF, PD and MSA. See text.

*6.2.3. Autonomic failure with Parkinson's disease (AFPD)* 

*6.2.4. Pure autonomic failure (PAF)* 

ganglionic abnormalities in DLB.

(cited from ref. 78)

PAF: Pure autonomic failure, PD: Parkinson's disease, GGL: ganglionic

MUPs.55

MUPs55.

MSA: multiple system atrophy, RU: residual urine volume, UD: underactive detrusor, AD: acontractile detrusor DSD: detrusor-sphincter dyssynergia

\*: International Cooperative Ataxia Rating Scale, walking capacities subscale (cited from ref. 74)

**Table 1.** Neurogenic sphincter EMG and clinical variables other than duration of illness.
