Somatic and Autonomic Function in Progressive Autonomic :Farlure and

Multiple System Atrophy Jeffrey Cohen, MD, Phillip Low, MD, Robert Fealey, MD, Sheldon Sheps, MD, and Nai-Siang Jiang, PhD

We studied 6 2 consecutive patients with progressive autonomic failure (PAF) cir multiple system atrophy (MSA) (26 PAF; 36 MSA). Patients were well matched in age (67 vs 66 years), duration (39 vs 36 months), and severity of autonomic failure (median values for PAF and MSA). Peripheral somatic neuropathy occurred in 2 patients with PAF and 7 patients with MSA. Postganglionic sudomotor and vasomotor functions were studied using the quantitative sudomotor axon reflex test and supine plasma norepinephrine. The extent and severity of autonomic failure were assessed by the thermoregulatory sweat test, by heart rate responses to deep breathing and the Valsalva maneuver, and by blood pressure recordings. Severe and widespread anhidrosis was found in both PAF and MSA patients. Postgangli- onic sudomotor failure occurred at the forearm in 50% each of PAF and MSA patients and at the foot in 69% and 66% of PAF and MSA patients, respectively. However, postganglionic sudomotor function was preserved in some patients with anhidrosis on thermoregulatory sweat test, indicating a preganglionic lesion. Vagal abnormalities were found in 77% and 81% of PAF and MSA patients. Supine plasma free norepinephrine values were significantly reduced in PAF (p < 0.001), but not in MSA, patients. Standing plasma norepinephrine values were reduced in both PAF (p < 0.001) and MSA (p < 0.001) patients. We conclude the following: (1) PAF is characterized by combined postganglionic sudomotor and adrenergic failure. (2) MSA is associated with a similar frequency of postganglionic sudomotor failure, but postganglionic adrenergic denemation was uncommon. (3) Preganglionic neuron is also involved in both disorders, but more severely in MSA. (4) Somatic neuropathy may occur.

Cohen J, Low P, Fealey R, Sheps S, Jiang N-S: Somatic and autonomic function in progressive autonomic failure and multiple system alrophy. Ann Neurol 22:692-699, 1987

Progressive autonomic failure (PAF, or idiopathic or- thostatic hypotension) and multiple system atrophy with autonomic failure (MSA, or Shy-Drager syn- drome) are characterized by severe orthostatic hy- potension (OH). The condition is designated MSA when there are associated clinical features of sttiatoni- gral or olivopontocerebellar involvement, and PAF when the central nervous system is uninvolved 11). In the past, we have found a marked depletion of pre- ganglionic sympathetic neurons in MSA, but not in PAF 121. There is recent pharmacological evidence of postganglionic adrenergic failure in PAF, leading to the hypothesis that PAF is a postganglionic disorder and MSA a preganglionic disorder [3-51. Somatic pe- ripheral nerve involvement in MSA has been sporadic- ally reported [2, 6-8).

We report an evaluation of 62 consecutive patients studied at the Mayo Autonomic Laboratories. We sought evidence of postganglionic sudomotor and ad- renergic failure using the combined results of the ther- moregulatory and quantitative sudomotor axon reflex

tests and plasma norepinephrine (NE) studies. A high prevalence of postganglionic sudomotor failure was found in both PAIF and MSA. In contrast, postgangli- onic adrenergic failure was frequently present in PAF but uncommon in MSA. Somatic neuropathy may oc- cur, especially in M A , and was usually mild.

Methods Strategy By performing the thermoregulatory sweat test (TST) [9,10] and the quantitative sudomotor axon reflex test (Q-SART) [lo] on the same pati.ents, we were able to determine (1) the distribution of anhiclrosis (TST data); (2) the presence of postganglionic failure. (Q-SART data); and (3) preganglionic involvement (anhidrosis by TST with a normal Q-SART). Supine NE measurements provided an index of postgangli- onic adrenergic failure [ 5 ] . Heart period responses to deep breathing and the Valsalva maneuver and blood pressure (BP), supine and after standing for 3 minutes, were also measured. These additional tests of parasympathetic and sympathetic function helped to provide an additional index of the extent and severity of autonomic failure.

From the Mayo Foundation, Rochester, MN 55905. Received Jan 23, 1987, and in revised form Apr 24 and May 27. Accepted for publication June 5, 1987.

Address correspondencc: to Dr Low.

692 Copyright 0 1987 by the American Neurological Association

Patients We studied all patients who were referred to the Autonomic Reflex Laboratory from March 1983 to February 1986. The diagnosis of autonomic failure required the presence of or- thostatic syncope or presyncope and sphygmomanometric confirmation of OH (orthostatic reduction in systolic BP by ;3, 30 mm Hg or mean arterial pressure of > 20 mm Hg). When striatonigral or olivopontocerebellar involvement was confirmed on neurological examination, the diagnosis of MSA was made. For the purposes of this study, a positive family history of a similar disorder was exclusionary.

Patients without central nervous system involvement were classified as PAF, provided there was no other identifiable cause for autonomic failure.

The diagnosis of peripheral neuropathy was made when at least two of the following five criteria were satisfied: neuropathic symptoms of numbness or paresthesias; distal weakness or muscle wasting; distal hyporeffexia or areflexia; electrophysiological changes of neuropathy; sural nerve biopsy findings of neuropathy.

Grading of autonomic failure followed these criteria:

1. Mild: Despite orthostatic symptoms and BP drop, the patient was able to maintain activities of daily living with- out medications or a body support stocking.

2. Moderate: The patient required medications andor a body support stocking to maintain activities of daily liv- ing.

3. Severe: The patient was severely limited in spite of medi- cations and could not remain on his or her feet for an hour at a time.

BP and Heart Rate Recordings BP was recorded using an automated BP and heart rate (HR) system. BP and H R were recorded supine and with the pa- tient tilted 80 degrees for 3 and 5 minutes.

Heart Rate Response to Deep Breathing and Valsalva Maneuver HR response was measured as previously described { 117. In brief, the HR variation to deep breathing at a rate of 6 breaths per minute was computer analyzed. The Valsalva ratio was calculated as the maximum H R generated by the Valsalva maneuver divided by the minimum HR occurring within 30 seconds of maximum H R {11-13}. The subject maintained a column of mercury at 40 to 50 mm for 15 seconds. The patient data were compared against age- matched values based on 72 control subjects aged 10 to 80 years. Patient response was reported as reduced when it fell below the first percentile.

Q-SART The methodology for the Q-SART has been previously de- scribed 1141. One population of eccrine sweat glands was stimulated by the iontophoresis of 10% acetylcholine, and the evoked sweat response was dynamically recorded from a second population of sweat glands. Minor recent modifica- tions include an increase in the size of the stimulus well and the use of the Demco Sudorometer (Demco Research and Development, Lansing, MI). The patient's response was re-

ported as reduced if it fell below the fifth percentile of con- trol values based on 100 controls (aged 10-90 years). Since control values varied by sex {14], the patient data were ex- pressed as a percentile of sex-matched control values. In early experiments, acetylcholine was also injected intrader- mally to ascertain if the reduced response was due to im- paired access to sweat glands in chronically anhidrotic pa- tients. There was excellent concordance between intradermal and iontophoresed responses, so subsequent experiments were performed with iontophoresis alone.

TST The TST done in the Mayo Thermoregulatory Laboratory was a modification of the Guttmann quinizarin sweat test 19, 10, 151: Unclothed subjects lay supine on a cart, and the exposed body surface (exclusive of eyes, nose, mouth, and genitalia) was covered with an even layer of indicator powder containing a mixture of alizarin monosulfonate, sodium car- bonate, and corn starch {lo}. The patient's initial oral tem- perature was recorded. The patient (including his or her head) was then enclosed in a cabinet (air temperature was between 44 and 50"C, relative humidity was between 40 and 50%, and overhead infrared heaters maintained the skin temperature between 38.5 and 39.5"C) for approximately 30 minutes. The patient's oral temperature was required to rise to 37.5"C or by 1.O"C above baseline (whichever yielded the higher temperature). The patient's sweat distribution (dark purple in sweating areas, light orange in anhidrotic areas) was observed and a human anatomical drawing was accurately shaded in to reflect the observation. The area of body surface anhidrosis was determined directly by planimetry (LASICO model 1252 M, Los Angeles Scientific Instrument Company, Los Angeles, CA, resolution 0.005 cm2).

The percent of body surface anhidrosis was computed ac- cording to the following formula:

measured area of anhidrosis 17.57 cm2

TST% =

The sweat distribution of patients was compared with results obtained from 35 healthy control subjects aged 20 to 75 years. All subjects were free of anticholinergic and gan- glionic blocking medications and were tested during the same part of the day.

Plasma Catecholamines Plasma catecholamines were measured using high-perfor- mance liquid chromatography with electrochemical detection {161 in 20 control subjects aged 20 to 65 years (9 males and 11 females). Catecholamines of patients with MSA who had taken levodopa-containing medications within the previous 2 weeks were excluded from the analysis.

Statistics All statistical calculations were carried out on a Hewlett- Packard HP9845T computer with data manipulation and ba- sic statistics, regression, ANOVA, and general statistics pro- grams. Linear regression was done using the regression program, and the slope was considered significant when p was less than 0.05 using a 2-tailed t test. For group data,

Cohen et al: Function in PAF and MSA 693

Table 1. Clinical ProfIeJ

Variable PAF MSA

Number 26 36 Sex (FIM) 1311 3 2 l / l 5

Duration (months)" 39 (4-128) 36 (6-84) Age (yearsy 67 (51-80) 66 (50-80)

Severity (1-3)" 2 (1-3) 2 (1-3) Number with neuropathy 2 7

"Median (5th-95th percentile). PAF = progressive autonomic failure; MSA = multiple system atrophy.

the distribution was nongaussian; nonparametric statistics (Mann-Whitney U test) were used.

Results Clinical Features Sixty-two patients satisfied our criteria for autonomic failure. Twenty-six were considered to have PAF and 36 to have MSA. The two groups of patients were well matched as to age and clinical severity of their auto- nomic failure, and with autonomic testing (Table 1). The duration of autonomic symptoms was slightly longer in PAF than in MSA patients but this difference was not statistically significant.

Clinical neuropathy occurred in 2 of 26 patients with PAF and 7 of 36 patients with MSA. The neurop- athy was a mild or asymptomatic distal sensorimotor polyneuropathy. Electrophysiological studies were un- dertaken when distal weakness, hypo- or areflexia, or hypesthesias were present. In all cases, motor and sen- sory studies were done on upper and lower extremity nerves, and motor unit potentials were routinely studied. The abnormalities were relatively mild, con- sisting of distal motor unit changes of chronic partial denervation, reduction of lower extremity compound motor andor sensory action potentials, and relatively preserved conduction velocities without significant dis- persion.

Electrophysiological studies were done on 7 of 9 patients with neuropathy. Surd nerve biopsy was done on 1 patient and showed a mild reduction in myelin- ated fiber density. Nine percent of teased single nerve fibers showed segmental remyelination, and 2% showed axonal degeneration. Another patient had bi- lateral Adie's pupils and electrophysiological and clini- cal evidence of a mild neuropathy, but his dominant disabilities were his autonomic failure and striatonigral degeneration. Both of these patients had signs of neuropathic involvement on examination. One patient with PAF had absent knee and ankle jerks and distal impairment to pinprick perception. The other patient with PAF had reduced ankle jerks and distal sensory loss. Only 2 patients had sensory symptoms which con- sisted of distal numbness and tingling.

Table 2. Summay of Autonomic Results for Progressive Autonomic Failure (PAF) and Multiple System Atrophy (MSA) Patients"

Variable PAF MSA

Q-SART (forearm) 6.5 (1-82) 6 (1-99) 13/26b 18/35b

18/26b 23135b Q-SART (foot) 1(1-33) 1 (1-60)

Valsalva ratio' 21/23 24/26 HR response to deep 17/22 24/27

Both VR and DB' 17/22 2 1/26 breathing'

MAP supine 107 (84-121) 108 (80-130) MAP standing 69 (47-94) 75 (47-89) HR supine 76 (59-90) 82 (64-98) HR standing 90 (65-115) 94 (70-111) TST percentd 91 (15-100) 97 (40-100) Norepinephrine supine 89 (27-379) 183 (47-433)

Norepinephrine standing 166 (31-759) 303 (91-830)

Norepinephrine increment 73 (0-100) 118 (18-315)

(pdml)

(pdml)

(pdml)

"Values are expressed as the median value (5th to 95th percentile). bNumber of patients below 5th percentile. 'Number of patients below 1st percentile. dPercent of total body anhidrosis on thermoregulatory sweat test. Q-SART = quantitative sudomotor axon reflex test; HR = heart rate; VR = Valsalva ratio; DB = deep breathing; MAP = mean arterial pressure; TST = thermoregulatory sweat test.

Sympathetic Sudomotor Function Widespread TST abnormalities were seen in most pa- tients (Table 2). Median values (and 5th to 95th per- centile) of body surface area anhidrosis were 91% (15-100%> and 97% (40-100%) for PAF and MSA patients, respectively. Differences were not statistically significant. Global anhidrosis with variable preserva- tion in acral parts was usual for both PAF and MSA patients (Fig 1). Not infrequently, striking regional deficits with a preserved segment of sweating were seen in PAF patients.

Q-SART responses were reduced or absent in the distal forearm of 50% of both PAF and MSA patients (Fig 2; see Table 2 ) . The frequencies of reduced or absent responses at the proximal foot (Fig 3) of PAF and MSA patients were almost identical (69% and 6696, respectively). The differences between the two conditions were not signhcantly different.

There was also evidence of preganglionic failure in both conditions. We examined a subset of PAF and MSA patients with anhidrosis on TST involving the Q- SART recording sites. The anhidrosis was designated as preganglionic or postganglionic depending on whether the Q-SART response was normal or absent. Based on studies of more than 100 patients with pe- ripheral neuropathy, we have confirmed the concor-

694 Annals of Neurology Vol 22 N o 6 December 1987

,?

w PA F

MSA PAF

m u Sweating No sweating

Fig 1 . Charactmitic thewnoregulatory sweating abnormalities in patients with progressive autonomic failure (PAF) or multiple system atrophy with autonomic failure (MSA, Shy-Drager syn- drome).

60 70 F

<5 5-25 26-50 51-75 76-99

Forearm Q-SART response as percentile

Fig 2. Quantitative sudomotor axon rejex test (Q-SART) re- sults, expessed as a percentile of control data, for the forearm in patients with progressive autonomic failure (PAF) or multiple system atrophy (MSA).

70

60

50 - m c 940 - 0

PAF (n=26) MSA (n-35)

$ 3 0

20

10

0 <5 5-25 26-50 51-75 76-99

Foot Q-SART response as percentile

Fig 3 . Quantitative sudomotor axon rejex test (Q-SART) re- sults, expressed as a percentile of control data, for the foot in patients with progressive autonomic failure (PAF) or multiple system atrophy (MSA).

Supine plasma norepinephrine, pg/rnl

Fig 4. Supine plasma norepinephrine in control subjects and in patients with progressive autonomic failure (PAF) or nult$le system atrophy (MSA).

dance berween TST and Q-SART results in post- ganglionic sudomotor failure 114, 171; these tests have similar sensitivities. Of the 10 PAF patients with TST anhidrosis involving forearm andor foot, 6 had combined pre- and postganglionic anhidrosis, in 3 the lesion was preganglionic, and in 1 the lesion was postganglionic. Of the 18 MSA patients with TST anhidrosis involving forearm andor foot, 4 had com- bined pre- and postganglionic anhidrosis, in 10 the lesion was preganglionic, and in 4 the lesion was post- ganglionic.

Sympathetic Adrenergic Function Supine plasma NE (Fig 4; Table 3) was significantly reduced (p < 0.001) in PAF (median value, 89 pg/ml), but not in MSA (median value, 183 pglml), patients when compared with controls (median value, 266 pgl ml). Values were significantly lower in PAF than in MSA patients (p = 0.029).

Standing plasma NE in both PAF and MSA patients was significantly reduced (p < 0.001 for both) below control values (Fig 5; see Table 3). The median incre- ment (see Fig 5 and Table 3) in NE of 73 pdml in PAF and 118 pg/ml in MSA patients was significantly re- duced (p < 0.001 for both PAF and MSA) when com- pared with the control increment (286 pg/ml).

Heart Period Recordings Heart period recordings (see Table 2) were reduced for Valsalva ratio and deep breathing in 91% and 77% of patients with PAF and in a similar percentage of MSA patients (92% and 89%, respectively). Seventy- seven percent of PAF patients and 81% of MSA pa- tients had both Valsalva ratio and deep breathing re- sponses that were abnormal.

Although a small percentage of patients had normal responses to one or two tests, the screening battery of tests used enabled us to diagnose widespread auto- nomic failure in all patients accepted as PAF or MSA.

Cohen et al: Function in PAF and MSA 695

Table 3. Plasma Free Norepinephrine (NEI in Control Subjects and Patients wi th Progressive Autonomic Failure or Multitlc System Atrophy wi th Autonomic FailurP"

Norepinephrine Control PAF MSA

Supineb 266 (1 11-5 18) 89 (27-379) 183 (47-433) Standing' 577 (322-1079) 166 (31-759) 303 (91-830)

286 (118-629) 73 (0-200) 118 (l8-3lr' Incrementd

"Values given in pg/ml; median value (5th to 95th percentile). bControl vs PAF: p < 0.001; control vs MSA: NS; PAF vs MSA: p = 0.029. 'Control vs PAF p < 0.001; control vs MSA: p < 0.001; PAF vs MSA: p = 0.024. dConrol vs PAF: p < 0.001; control vs MSA: p < 0,001; PAF vs MSA: NS.

30

b Z20 8 0

10

n n Norepinephrine increment (pg/ml) on standing

Fig 5 . Plarma norepinephrine increment in control subjects and in patients wi th progressive autonomic failure (PAF) or multiple system atrophy (MSA).

Regression of Q-SART and Supine NE with Other Tests We regressed the indexes of postganglionic sudomotor functions (Q-SART) with duration, severity, supine NE, and percent TST anhidrosis, indexes of the sever- ity and extent of autonomic failure (Table 4). For pa- tients with MSA (see Table 4), Q-SART regressed significantly against severity (p < 0.005) and supine NE (p < 0.01). For patients with,PAF, Q-SART re- gressed significantly against severity f$ < 0.05) and TST percent (p < 0.0025).

Discussion The 26 patients with PAF and 36 patients with MSA were well matched in age, duration, and severity of autonomic failure. They all had severe generalized au- tonomic failure involving the sympathetic nervous sys- tem as indicated by their severe OH, plasma NE stud- ies, TST, and Q-SART studies. The HR responses to deep breathing and the Valsalva maneuver demon- strated that the two groups of patients had a similar degree of parasympathetic failure.

Progressive Autonomic Failure Our data present a new finding that postganglionic sudomotor failure occurs in most patients with PAF (69% in the foot and 50% in the forearm). The re- duced plasma NE suggests postganglionic adrenergic

failure. The NE data agree closely with results from an earlier published study of PAF and MSA 19) and confirm earlier reports of postganglionic adrenergic failure in PAF {4, 5 ) in smaller patient populations. Plasma NE results from the spillover of NE from sym- pathetic postganglionic nerve terminals; therefore, the supine value is a net index of sympathetic activity [4, 51 and NE secretion and clearance E18, 191. Ad- renergic failure has more directly been demonstrated in the postganglionic neuron [201 and in terminals { 3 , 211.

Although there is clear evidence for postganglionic sympathetic failure in PAF, we think the disorder should be considered a combined preganglionic and postganglionic autonomic failure rather than a pure postganglionic disorder for the following reasons: (1) There is morphological evidence of preganglionic fail- ure [ 2 , 22). ( 2 ) There is evidence of preganglionic sudomotor failure in the present study. ( 3 ) There is a significant subset of patients with OH associated with normal supine NE that fails to rise on standing, a pat- tern suggestive of failure of preganglionic sympathetic activation [51. (4) Patients with PAF seem to have less postganglionic sympathetic failure than neuropathic patients (where the sympathetic lesion is known to be postganglionic), who have a similar degree of ortho- statism and anhidrosis (Low and Fealey, unpublished observations). These clearly postganglionic neuropa- thies, eg., acute panautonomic neuropathy { 2 31, have uniformly absent Q-SARTs and a severe reduction in supine NE values in contrast to PAF. (5) In contrast, some patients with PAF have generalized anhidrosis and entirely normal postganglionic sympathetic func- tion. Ninety percent of patients with PAF had some preganglionic sudomotor failure, although the preva- lence of pure preganglionic anhidrosis was lower in PAF (22%) than in MSA (56%) patients.

Some of the patients with PAF were elderly, and OH of the elderly has been suggested by some investi- gators as unrelated to autonomic failure. Therefore, the criticism could be made that the normal supine NE values (seen in about a third of our patients with PAF) may be attributable to this group (elderly patients with

696 Annals of Neurology Vol 22 No 6 December 1987

Table 4. Regression Analyses on Progressive Autonomic Failure and Multiple System Atrophy Patients

Independent Variable vs Dependent Variable Equation R T Significance

PROGRESSIVE AUTONOMIC FAILURE

Q-SART vs duration Y = 60.05 - 0.238X 0.117 0.58 NS p-SART vs severity Y = 2.228 - 0.01ox 0.354 1.85 < 0.05 Q-SART vs TST percent Y = 84.607 - 0.667X 0.558 3.29 < 0.0025 Q-SART vs NE (supine) Y = 164.486 - 0.557X 0.083 0.34 NS

MULTIPLE SYSTEM ATROPHY

Q-SART vs duration Y = 43.70 + 0.223X 0.084 0.49 NS Q-SART vs severity Y = 2.330 - 0.011X 0.335 2.04 < 0.005 Q-SART vs TST percent Y = 86.472 - 0.004X 0.004 0.03 NS Q-SART vs NE (supine) Y = 179.916 + 5.448X 0.598 2.47 < 0.01

Q-SART = quantitative sudomotor axon reflex test; TST = thermoregulatory sweating test; NE = norepinephrine.

OH). OH in the elderly is uncommon and mild 1241 and its characteristics include a relatively mild and asymptomatic OH associated with normal heart period “to standing” and normal supine and standing plasma NE values 1253. However, many elderly patients de- velop OH when appropriately stressed 1261. Auto- nomic impairment affects most autonomic systems with increasing age. Subjects may have problems with thermoregulatory responsiveness, impotence, and im- pairment of gastrointestinal, urinary, lacrimal, and pupillary functions 127, 281. Our own data on the 5 to 8% attrition of intermediolateral column cell counts 1291, especially when validated in the same cases with preganglionic axon counts in the ventral spinal root 1301 and rami communicantes 1311 at the same level, provide strong morphological evidence for autonomic neuronal attrition.

To examine if there was a segregation of elderly patients with OH but without autonomic failure, this subset of patients was analyzed. When supine NE level was regressed against age, a significant regression was unobtainable. When patients above the age of 70 years were separately considered, we were unable to obtain a significant difference from patients under the age of 70 years. When other parameters were considered (TST percent, Q-SART, or heart period responses to deep breathing and the Valsalva maneuver), the older subjects were still nonsegregable. In this report, the older subjects have autonomic failure involving cardiac autonomic control, sudomotor control, NE levels, as well as symptomatic OH.

Clinical criteria for the diagnosis of PAF are imper- fect, and much of the controversy about the site and purity of the lesion possibly relates to the heterogene- ity of what is designated PAF. The entity probably comprises patient subsets who have OH in common. Some patients appear to have a predominantly post-

ganglionic adrenergic disorder ab initio C4, 51. Others have a combined postganglionic and preganglionic le- sion, as suggested by this study. A small subset begins with PAF and progresses to MSA 111. Some patients develop diffuse autonomic failure with increasing age 1271, yet others develop OH without obvious auto- nomic failure 1251.

Multiple System Atrophy Evidence for MSA being a primarily preganglionic dis- order is compelling. Apart from its association with involvement of other central nervous structures (espe- cially striatonigral and olivopontocerebellar), MSA is characterized by a normal resting NE. The relatively normal resting NE of MSA is in contrast to the marked reduction noted in disorders such as an acute panautonomic neuropathy 12 31 or diabetic autonomic neuropathy 1321; again, this suggests a preganglionic site for MSA. The significantly reduced NE response to standing in MSA is thought to indicate a failure to activate central sympathetic pathways 151. These findings are consonant with published reports of a marked reduction in preganglionic sympathetic neu- rons 12, 22, 33-373, including morphometric studies 12, 22, 33-35, 371. The splanchnic autonomic outflow is of crucial importance in the maintenance of orthos- tatic normotension, and the attrition of 5 to 8% of neurons per decade 130, 31, 381 may be responsible for the age-related OH 1127. Autonomic neuropathies with prominent OH are associated with loss of inter- mediolateral column neurons 12, 12, 29), and OH cor- relates with the attrition of approximately 50% of in- termediolateral column neurons 112).

Our patients with MSA had less postganglionic fail- ure than patients with PAF, who were largely confined to postganglionic sudomotor failure with relative spar- ing of postganglionic adrenergic function. The subset

Cohen et al: Function in PAF and MSA 697

of patients permitting precise localization of anhidrosis (combined Q-SART and TST studies) demonstrated a higher percent of pure preganglionic lesions in MSA (56%) than in PAF (22%). The high percent of abnor- mal Q-SART results suggests that sudomotor failure precedes vasomotor failure, a finding that is consistent with the clinical impression that TST may be a sensi- tive screen of generalized autonomic failure, since the sudomotor abnormalities often antedate OH.

Although MSA patients appear to be a relatively homogeneous group with only a minor subset exhib- iting postganglionic adrenergic failure, the latter has been well demonstrated pharmacologically 1391 and on catecholamine fluorescence and electron microscopy of peripheral veins [2 11.

Our data lend some support to the hypothesis that the postganglionic sympathetic sudomotor failure in MSA is due to a transsynaptic effect. If the mechanism is transsynaptic, the degree of postganglionic failure should vary with the severity of autonomic failure. In MSA, there is a significant regression of Q-SART values with clinical severity, percent of anhidrosis on TST, and supine plasma NE, which supports this hy- pothesis. The greater percent of postganglionic failure in the foot than the forearm suggests a possible length- dependent transsynaptic mechanism. The insignificant regression with duration is not surprising, since these cases are chronic in nature.

Although there is a good statistical separation be- tween groups of patients with PAF and MSA using supine plasma NE, there is also considerable overlap. Therefore, supine plasma NE is of limited value in separating the individual patient with PAF from the one with MSA. Patients with PAF often have normal supine plasma NE levels (see Fig 4). Moreover, some patients with MSA have reduced NE (see Fig 4) (401. The most reliable way to diagnose MSA is still the demonstration of central nervous system involvement. Finally, the occasional patient may progress from ap- parent PAF to MSA 111.

Peripheral neuropathy may occur in both conditions (more common in MSA) as was demonstrated in our 2 patients with PAF and 7 patients with MSA. This confirms earlier reports of somatic involvement at the level of the roots [ 2 ] or peripheral nerve [6, 81. How- ever, the neuropathy is usually mild or asymptomatic.

Supported in part by grants from NINCDS (R01 NS22352, NS14304), MDA, and Mayo and Mogg Funds.

References 1. Thomas JE, Schirger A. Idiopathic orthostatic hypotension. A

study of its natural history in 57 neurologically affected persons. Arch Neurol 1970;22:289-293

2. Low PA, Thomas JE, Dyck PJ. The splanchnic autonomic

outflow in Shy-Drager syndrome and idiopathic orthostatic hy- potension. Ann Neurol 1978;4:5 11-5 14

3. Kontos HA, Richardson DW, Norvell JE. Norepinephrine de- pletion in idiopathic orthostatic hypotension. Ann Intern Med 1975;82:336-341

4. Polinsky RJ, Kopin IJ, Ebert MH, Weise V. Pharmacologic distinction of different orthostatic hypotension syndromes. Neurology 198 1 ;3 1 : 1-7

5. Ziegler MQ, Lake CR, Kopin IJ. The sympathetic nervous sys- tem defect in primary orthostatic hyotension. N Engl J Med

6. Galassi G, Nemni R, Baraldi A, et al. Peripheral neuropathy in multiple system atrophy with autonomic failure. Neurology

7. Low PA, Stevens JC. Peripheral neuropathies including auto- nomic neuropathies. In: Spinel JA Jr, ed. Clinical medicine. New York: Harper & Row, 1985

8. Toghi H, Tabuchi M, Tomonaga M, Izumiyama N. Selective loss of small myelinated and unmyelinated fibers in Shy-Drager syndrome. Acta Neuropathol (Berl) 1982;57:282-286

9. Fealey RD, Schirger A, Thomas JE. Orthostatic hypotension. In: Spinel JA, ed. Clinical medicine. New York: Harper & Row, 1985

10. Low PA, Walsh JC, Huang CY, McLeod JG. The sympathetic nervous system in diabetic neuropathy. A clinical and patho- logical study. Brain 1975;98:341-356

11. Low PA, Zimmerman BR, Dyck PJ. Comparison of distal sym- pathetic with vagal function in diabetic neuropathy. Muscle Nerve 1986;9:592-596

12. Low PA. Quantitation of autonomic function. In: Dyck PJ, Thomas PK, Lambert EH, Bunge R, eds. Peripheral neuropa- thy. 2nd ed. Philadelphia: Saunders, 1984

13. Wheeler T, Watkins PJ. Cardiac denervation in diabetes. Br Med J 1973;4:584-586

14. Low PA, Caskey PE, Tuck RR, et al. Quantitative sudomotor axon reflex test in normal and neuropathic subjects. Ann Neurol 1983;14:573-580

15. Guttmann L. The management of the quinizarin sweat test (QST). Postgrad Med J 1947;23:353-366

16. Jiang N-S, Machcek D. Measurement of catecholamines by liq- uid chromatography with amperometric detection. In: Parvez H, et al, eds. Progress in HPLC. Vol2. Utrecht, Holland: VNU Science Press, 1987;395-424

17. Low PA, Fealey RD. Sudomotor neuropathy. In: Dyck PJ, Thomas PK, Asbury AK, Winegrad AI, eds. Diabetic neuropa- thy. Philadelphia: Saunders, 1986

18. Esler M, Jackson G, Kelleher D, et al. Norepinephrine kinetics in patients with idiopathic autonomic insufficiency. Circ Res

19. Polinsky RJ, Goldstein DS, Brown RT, et al. Decreased sympa- thetic neuronal uptake in idiopathic orthostatic hypotension. Ann Neurol 1985;18:48-53

20. Petito CK, Black IB. Ultrastructural and biochemistry of sympa- thetic ganglia in idiopathic orthostatic hypotension. Ann Neurol

21. Bannister R, Crowe R, Eames R, Burnstock G. Adrenergic innervation in autonomic failure. Neurology 1981;31:150-156

22. Johnson RH, Lee G de J, Oppenheimer DR, Spalding JMK. Autonomic failure with orthostatic hypotension due to inter- mediolateral column degeneration. A report of two cases with autopsies. Q J Med 1966;35:276-292

23. Low PA, Dyck PJ, Lambert EH, et al. Acute panautonomic neuropathy. Ann Neurol 1983;13:412-417

24. Rodstein M, Zeman FD. Postural blood pressure changes in the elderly. J Chronic Dis 1957;6:681-688

25. Robinson BJ, Johnson RH, h b i e DG, Palmer KT. Do elderly

1977;296:293-297

1982;32:1116-1121

1980;46(~~ppl 1): 147-148

1978;4:6-10

698 Annals of Neurology Vol 22 No 6 December 1987

patients with an excessive fall in blood pressure on standing have evidence of autonomic failure? Clin Sci 1983;64:587-591

26. Lipsitz LA, Nuquist RP, Wei JY, Rowe JW. Postprandial re- duction in blood pressure in the elderly. N Engl J Med 1983;

27. Collins KJ. Autonomic failure in the elderly. In: Bannister R, ed. Autonomic failure. New York: Oxford U Press, 1983;489- 507

28. Foster KG, Ellis FP, Dor6 C, et al. Sweat responses in the aged. Age Ageing 1976;5:91-101

29. Low PA, Dyck PJ, Okazaki H , et al. The splanchnic autonomic outflow in amyloid neuropathy and Tangier disease. Ann Neurol 1981;3 ~461-463

30. Low PA, Dyck PJ. Splanchnic preganghonic neurons in man. 11. Preganglionic ventral root fibers. Acta Neuropathol (Berl)

31. Low PA, Dyck PJ. Splanchnic preganglonic neurons in man. 111. Morphometry of myelinated fibers of rami communicantes. J Neuropathol Exp Neurol 1978;37:734-740

32. Christensen NJ. Plasma catecholamines in long-term diabetics with and without neuropathy and in hypophysectomized sub- jects. J Clin Invest 1972;51:779-787

309~81-83

1977;40:2 19-226

33. Bannister R, Oppenheimer DR. Degenerative diseases of the nervous system associated with autonomic failure. Brain

34. Kennedy PGE, Duchen LW. A quantitative study of inter- mediolateral column cells in motor neuron disease and the Shy- Drager syndrome. J Neurol Neurosurg Psychiatry 1985;

35. Oppenheimer DR. Lateral horn cells in progressive autonomic failure. J Neurol Sci 1980;46:393-404

36. Shy GM, Drager GA. A neurologic syndrome associated with orthostatic hypotension-a clinical-pathologic study. Arch Neurol 1960;2:511-527

37. Sung JH, Mastri AR, Segal E Pathology of Shy-Drager syn- drome. J Neuropathol Exp Neurol 1979;38:353-368

38. Low PA, Okazaki H, Dyck PJ. Splanchnic preganglionic neurons in man. I. Morphometry of preganglionic cytons. Acta Neuropathol (Berl) 1977;40: 5 5-6 1

39. Steiner JA, Low PA, Huang CY, et al. L-dopa and the Shy- Drager syndrome. Med J Aust 1974;2:133-136

40. Bannister R, Sever P, Gross M. Cardiovascular reflexes and biochemical responses in progressive autonomic failure. Brain

1972;95:457-474

48:1103-1106

1977;100:327-344

Cohen et ak Function in PAF and MSA 699