Cerebrospinal Fluid as a Reflector of Central Cholinergic and Amino Acid Neurotransmitter Activity in Cerebellar Ataxia Bala V. Manyam, MD; Ezio
Giacobini, PhD; Thomas N. Ferraro, PhD; Theodore A. Hare, PhD
\s=b\ Cerebrospinal fluid (CSF) amino acid neurotransmitters, related compounds, and their precursors, choline levels, and acetycholinesterase activity were measured in the CSF of patients with cerebellar ataxia during a randomized, double\x=req-\ blind, crossover, placebo-controlled clinical trial of physostigmine salicylate. The CSF \g=g\-aminobutyricacid, methionine, and choline levels, adjusted for age, were significantly lower in patients with cerebellar ataxia compared with controls. Physostig-
mine selectively reduced the level of CSF isoleucine and elevated the levels of phosphoethanolamine. No change occurred in CSF acetylcholinesterase activity and in the levels of plasma amino compounds in patients with cerebellar ataxia when compared with controls. Median ataxia scores did not statistically differ between placebo and physostigmine nor did functional improvement occur in any of the patients. (Arch Neurol. 1990;47:1194-1199)
í^ erebellar ataxia represents
a diffuse group of disorders for which both inherited and noninherited forms have been described. Age at onset is variable but usually occurs in middle age. Studies on the cerebrospinal fluid (CSF) neurochemistry of noninherited
Accepted for publication March 16, 1990. From the Division of Neurology (Dr Manyam) and Department of Pharmacology (Dr Giacobini), Southern Illinois University School of Medicine, Springfield, and the Departments of Pharmacology (Drs Ferraro and Hare) and Neurology (Dr Ferraro), Thomas Jefferson University, Philadelphia, Pa. Reprint requests to Division of Neurology, D411, Southern Illinois University School of Medicine, PO Box 19230, Springfield, IL 62794-9230 (Dr Manyam).
cerebellar disorders
are
scarce1-2; how¬
ever, abnormalities of amino acid
neu¬
rotransmitter systems have been
ported in postmortem brains of
re¬
pa¬
tients who died with inherited forms of cerebellar dysfunction.35 Deficiency of glutamate dehydroge¬ nase has been reported in a subset of patients with olivopontocerebellar atrophy.68 Glutamate dehydrogenase forms part of a complex enzymatic process for transferring amino groups between a-keto acids and their corre¬ sponding -amino acids. Furthermore, investigation of amino acids in cere¬ bellar disorders is particularly appro¬ priate since at least seven major types of cerebellar neurons are believed to use amino acids as their neuro¬ transmitters.913 -Aminobutyric acid (GABA) is believed to be the inhibitory neurotransmitter of Purkinje's cells9 and may also be used by the three in¬ hibitory interneurons in the cerebellar cortex—the basket and the stellate and Golgi's cells10—although it has been suggested that taurine rather than GABA may be the inhibitory neurotransmitter used by the stellate cells.11 Glutamate is believed to be the neurotransmitter used by the mossy and climbing fibers and by the granule cells with their parallel fibers,12 al¬ though evidence also exists to impli¬ cate aspartate as the neurotransmit¬ ter of parallel fibers.13 A study of amino acids in postmor¬ tem brain tissue from 10 patients (six different pedigrees) with dominantly inherited cerebellar disorders" re¬ vealed three biochemically different disorders despite clinical similarities. One disorder was characterized by
moderate reduction of aspartate and
glutamate levels in cerebellar cortex; a second disorder was characterized by reductions of aspartate and glutamate levels in other brain regions in addi¬
tion to cerebellar cortex; and a third disorder was characterized by normal aspartate and glutamate concentra¬ tions in all brain regions examined. The GABA content in cerebellar cortex and dentate nucleus was reduced in some patients with each disorder, whereas cerebellar taurine was nor¬ mal in all patients. Increased GABA receptor binding in the cerebellar cor¬ tex of patients who died with cerebel¬ lar ataxias has been documented"1 and this finding is consistent with de¬ creased GABA levels in the dentate nucleus.4 While the full CSF amino acid profile for cerebellar ataxia has not been previously reported, it has been documented that CSF GABA is significantly reduced in this disorder.1·2 Several inherited ataxias are asso¬ ciated with defects of pyruvate oxida¬ tion, and such defects are known to re¬ sult in the inhibition of acetylcholine (ACh) synthesis in the brain.15 The ac¬ tivity of choline acetyltransferase, the enzyme that synthesizes ACh from choline, and ACh, are markedly re¬ duced in the brain of patients who die with olivopontocerebellar atrophy.16 Physostigmine salicylate has a revers¬ ible inhibitory effect on cholinesterase; as a consequence, it enhances cholin¬ ergic transmission at synapses. The drug is readily absorbed from the gas¬ trointestinal tract, easily crosses the blood-brain barrier, and acts on the central nervous system." In a doubleblind, randomized, clinical trial of
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
physostigmine, we evaluated possible alterations of cholinergic function and
Table 1.—Physostigmine Treatment of Cerebellar Ataxia: Clinical
amino acid neurotransmitters in pa¬ tients with cerebellar ataxia by mea¬ suring levels of CSF amino acid neuro¬ transmitters, related compounds, their precursors, and acetylcholinesterase
Patient
Summary Age, y Family history
(AChE) activity.
PATIENTS AND METHODS
seen on a
±
SD age of 53
±
cholinergic agonistic or antagonistic drugs. No dietary restrictions were im¬ posed. Informed consent was obtained from patients, and approval for the trial was granted by the Institutional Committee on Human Experimentation. To study the effect of physostigmine on any
CSF neurochemicals and cerebellar symp¬ toms, physostigmine was administered with the use of a placebo-control, doubleblind crossover design so that the CSF col¬ lected during the placebo phase would re¬ flect changes due to disease, and any addi¬ tional changes in CSF collected during the physostigmine phase would indicate changes due to treatment. Physostigmine salicylate (1 mg) or identical placebo (lac¬ tose) tablets were allotted to patients with ataxia who were chosen randomly for 12 weeks each. At the end of 12 weeks, physo¬ stigmine and placebo were crossed over for another 12 weeks. Either physostigmine or placebo was administered at a rate of one tablet three times each day during the first week, one tablet four times each day during the second week, one tablet six times each day during the third week, and then one tablet given once every 2 hours, except while asleep (eight times each day) from the fourth through the 12th week. Before the start of the trial, patients were videotaped and then retaped during the 12th week of both placebo and drug treatment. All video tape recording was done at approximately the same time of day. One hour before recording time, a dose of either physostig¬ mine (or placebo) was administered so as to induce a maximum therapeutic effect. Cerebellar function was assessed by semiquantitative scoring as follows: 0, absence of ataxia; 1, minimal or extreme normal; 2, mild; 3, moderate; 4, severe; and 5, inability to carry out the test.18 The total score for each observation was used for comparison. Reduction in score indicated improvement. Complete blood cell counts, liver and renal function test results, serum electrolyte de¬ terminations, and calcium and phosphorus
62
47
52
35
53 ± 11 + 0/6
Alcohol intake Skeletal deformities
0/6
Dementia
0/6 5/6
83
4/6 6/6 6/6
67 100 100
6/6
100
Intention tremor
Dysdladochoklnesia Truncal ataxia Gatt ataxia
computed tomographic (CT) scan.
Patients 2, 3, 4, and 6 gave a history of oc¬ casional alcohol intake, but other features of chronic alcoholism, such as dementia, peripheral neuropathy, and reversal of al¬ bumin-globulin ratios in the serum, were lacking. None of the patients were receiving
62
Optic atrophy Dysarthria Nystagmus Dysmetria, F-N Dysmetria, H-S
11 years (range, 35 to 62 years) participated in the trial. The details of clinical data are presented in Table 1. All patients were con¬ sidered to have sporadic cerebellar ataxia, except patient 1, in whom early olivoponto¬ cerebellar atrophy could not be excluded as an extensor plantar response with cerebel¬ lar atrophy without brain-stem atrophy as
Six men with a mean
Findings*
83
Muscle tone DTRs Plantar responses CT scan, cerebellar
i
1
Hyper
Hypo
Hypo
tt
il
li
atrophy
u li
ii
ii
17 4/6
0
67
*
Asterisk indicates unable to test; 0, negative or none; plus sign, present or positive; F-N, finger to nose test; H-S, heel to shin test; N, within normal limits; DTRs, deep tendon reflexes; Hyper, hyperactive; Hypo, hypoactlve; upward arrow, extensor response; downward arrow, flexor response; an arrow, each lower limb; and CT,
computed tomographic. tMean ± SD.
levels riod.
were
monitored
during the trial
pe¬
Lumbar Puncture
Cerebrospinal fluid was collected by lum¬ bar puncture at the end of 12 weeks of both placebo and physostigmine treatment phases. The CSF was collected between 8 and 9 am in all patients following overnight bed rest and fasting (water was allowed ad libitum to prevent dehydration) for 12 hours. In patients under treatment for cerebellar ataxia, either placebo or physo¬ stigmine salicylate (1 mg) was adminis¬ tered 1 hour before CSF collection. Lumbar puncture was performed in the standard manner. The details of CSF collection, prep¬ aration, and storage conditions have been described elsewhere.19 Cerebrospinal fluid for measurement of neurochemicals from both controls and the experimental group was from the same gradient (first 10 mL of CSF) and, thus, was not subject to gradient
Acetylcholinesterase retains enzyme activity in CSF after long-term storage at -70°C and even after repeated freezing and thawing.20 The reliability of freezing naive CSF before high-performance liquid chro¬ matography analysis for CSF amino com¬ pounds and of using the same gradient from effect.
the first 10 mL of CSF has been confirmed in our laboratory21 and others.22 Control CSF was obtained from 17 men without evidence of organic neurologic dis¬ ease or mental disorders. Of these, CSF samples from 10 men with a mean ± SD age of 48 ± 20 years were used to act as controls for amino compound studies, and samples from nine men with a mean ±SD age of 64 ± 9 years were controls for choline lev¬ els and AChE activity studies with an over¬ lap of two controls. This was done due to the limited availability of CSF from controls,
as
the CSF from the first 10 mL was used for
comparison between the controls and ex¬ perimental group to control gradient effect. Protein, glucose, sodium, potassium, chloride, calcium, and phosphorus values were measured in the clinical laboratory with an automated analyzer. The IgG level was
measured with radioimmunodiffusion. Amino
Compound Analysis
Cerebrospinal fluid and plasma amino were quantified with the use of a tri¬ ple-column, ion-exchange Chromatographie procedure with fluorometric detection.23 Cerebrospinal fluid and plasma proteins were precipitated by adding a one-third volume of 20% aqueous sulfosalicylic acid. Specimens were centrifuged at 2000¡7, and acids
the supernatant fluids were used for direct injection onto the high-performance liquid Chromatograph. Norleucine and 5-aminovaleric acid were used as internal stan¬ dards. The analytical procedure used par¬ allel operation of three distinct ionexchange columns for separation and mea¬ surement of acidic, neutral, and basic amino acids, respectively, with a recovery rate for an internal standard of 98% and a lower limit of detection of 5 pmol. The chromatography was run isocratically in each case with the use of lithium citrate buffers designed to provide overlapping elution profiles. All samples were run in duplicate, and the mean was used for calcu¬ lation.
Assay of AChE Activity The radiometrie method of Johnson and Russell24 was used to determine AChE ac¬ tivity with tritiated ACh iodide as a sub¬
(90 mCi/mmol, specific activity). Acetylcholinesterase activity was exstrate
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
Table 2.—CSF Amino
Compounds
pressed as micromoles of tritiated ACh io¬ dide hydrolyzed per hour per milliliter of
in Cerebellar Ataxia* Mean ±
SD,
nmol/mL
Patients With Cerebellar Ataxia ( 6) 0.051 ± 0.024
Controls Amino Compounds /3-Alanlne
0.078 ± 0.035
GABA
0.142 ± 0.035
Aspartate 1-Methylhlstidine 3-Methylhlstldine
0.275 ± 0.151
0.084 ± 0.012t 0.223 ± 0.159
1.24 ± 1.12
1.19 ± 0.736 1.20 ± 0.484
(n
=
10)
=
Homocarnosine
1.30 ± 1.01 1.84 ± 1.10
Tryptophan
2.75 ± 1.67
0.855 ± 0.421 2.76 ± 1.25
Methyl-lysine
2.83 ± 3.10
3.44 ± 2.62
Citrulline
2.88 ± 1.02
2.63 ± 0.591
tj-Amlnobutyrate
3.31
3.90 ± 1.34
1.08
Glutamate
3.13 ± 0.395 3.51 ± 0.729 3.76 ± 0.957
Phosphoethanolamine Methionlne
3.09 ± 0.609
2.37 ± 0.517t
Ornithine Isoleuclne Taurine
4.63 ± 0.664
5.29 ± 0.915
4.95 ± 1.18
5.99 ± 0.740 5.17 ± 1.75
Glycine
8.08 ± 4.26
7.15 ± 0.631
Asparagine
9.24 ± 1.93
9.08 ± 1.49 8.87 ± 1.48
Ethanolamlne
9.01
Phenylalanine Tyrosine
10.16 ± 2.62 10.73 ± 5.19
8.32 ± 2.36
Histidlne
12.06 ± 2.53 13.28 ± 2.76
12.13 ± 2.73 13.75 ± 1.87
Valine
16.69 ± 4.01
19.88 ± 2.58
Arginine Serine
22.25 ± 4.01 25.41 ± 4.79
25.65 ± 4.01
Lysine
25.69 ± 4.46
29.37 ± 2.74
Alanine Threonine Glutamine
30.44 ± 6.90 33.77 ± 4.18
Leuclne
9.58 ± 2.43
29.27 ± 4.81
34.48 ± 5.06 655.3 ± 63.6
311.5 ± 67.2
*CSF indicates cerebrospinal fluid; GABA, -aminobutyric acid. tBy analysis of covariance with age as covariate (P .0001).
Table 3.—Effect of
=
SD, nmol/mL
Phosphoethanolamine
Isoleucine
3.09 ± 0.61 3.64 ± 0.67
5.99 ± 0.74
.0072
.0209
Physostigmine salicylate Slgnlflcancet *CSF indicates cerebrospinal fluid. tBy paired Student's i test.
Controls Patients with cerebellar ataxia
Significance, P(CSF AChE)
0.28
0.75 ± 0.20 .2440t Physostigmine sallcylatet 6 0.74 ± 0.21 ,5678§ *CSF Indicates cerebrospinal fluid; AChE, acetylcholinesterase. tBy analysis of covariance with age as covariate. tDouble-blind crossover trial with final dose at 8 mg/d for 9 weeks. §By paired Student's t test. Placebo
•
ease
(1.36/0.03 mmol-mL-h).25 Choline Assay
For the assay of CSF choline, 1 mL of CSF mixed with 10 mL of IN formic acidacetone (ratio, 15:85) and centrifuged at 30000 for 20 minutes to remove protein. The levels of choline were measured in 15-uL aliquots of CSF by the radiometrie enzy¬ matic assay.26 Absolute levels of choline were calculated on the basis of standards (10,50, and 100 pmol). All samples were as¬ sayed in duplicate. was
Statistical Analysis
Computations were done on an IBM 4341
ance with age as covariate was used to con¬ trol the effect of age on CSF and plasma components between controls and patients with cerebellar ataxia. The paired t test was applied to compare the effect of physostig¬ mine between placebo and treated phases, as this was a crossover trial. On ataxia rat¬ ing scores, the Wilcoxon matched-pairs signed-rank test was used.
RESULTS
Table 4.—Cerebellar Ataxias: CSF AChE Activity Choline Levels* CSF AChE, mL µ per h
of 10_i mol/L. At this concentration, isoompa inhibited 90% BuChE CSF activity measured with tritiated butyrylcholine io¬ dide as a substrate. In CSF, BuChE activity represents approximately 5% of total cholinesterase activity, and there is no dif¬ ference, for example, in the ratio of AChE:BuChE for controls (1.35/0.061 mmol mL-h) vs patients with Alzheimer's dis¬
procedure27 was used to estimate means, SDs, and significance. Analysis of covari¬
on CSF Amino Compounds in Patients With Cerebellar Ataxias ( 6)*
Placebo
tetraisopropylpyrophosphoramide, Sigma Chemical Co, St Louis, Mo) of 10"3 to 10"" mol/L. Based on these experimental re¬ sults, to inhibit BuChE activity present in CSF, iso-ompa was used at a concentration
computer. The Statistical Analysis System
Physostigmine
Mean ±
CSF. To determine the proper concentra¬ tion of the inhibitor, inhibition experi¬ ments were performed on CSF, as well as on purified samples of butyrylcholinesterase (BuChE, serum) and AChE (erythrocytes), at decreasing concentration of inhibitors (BW284C51, Burroughs Wellcome Co, Re¬ search Triangle Park, NC, and iso-ompa,
CSF
Choline, nmol/mL
Significance, /»(CSF Choline)
2.97 ± 0.79
1.02 ± 0.29 0.96 ± 0.19§
.0003t 3825§
To study the alteration of CSF amino compounds in patients with cerebellar ataxia, CSF obtained during the placebo phase was compared with that of controls. During the placebo phase, patients were monitored closely for 12 weeks before CSF collection, and thus, the values could be considered to
represent
a more
homogeneous
popu¬
lation than that represented by ran¬ domly collected samples. To study the effect of physostigmine on CSF and plasma amino compounds, values gen¬ erated from samples collected at the end of 12 weeks of placebo and 12 weeks of physostigmine were compared. With the use of analysis of covariance, with age as covariate, a significant re-
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
cerebellar ataxia. Methionine derived from diet or other tissues crosses the blood-brain barrier, and differences in methionine levels in the various brain regions are not great.28 30 Methionine in the free amino acid pool of the brain appears to be used mainly for protein synthesis.31 Methionine has been shown to have a significant effect when administered to schizophrenic pa¬ tients, causing a marked aggravation of symptoms.32 While this establishes the importance of methionine in cen¬ tral nervous system function, its role in cerebellar function needs further
study. Physostigmine treatment selec¬ tively elevated the levels of phospho-
Effect of oral physostigmine treatment (dotted bars) on cerebellar ataxia. dicates improvement; open bars, placebo treatment.
duction ( .0001) of CSF GABA and methionine was found in patients with cerebellar ataxia when compared with controls (Table 2). Physostigmine se¬ lectively elevated the levels of phosphoethanolamine (P .0072) and re¬ duced the levels of isoleucine (P .0209) but had no significant ef¬ fect on levels of other CSF amino com¬ =
=
=
pounds (Table 3).
Mean ± SD CSF choline levels, ad¬
justed for age, were significantly lower for patients with cerebellar ataxia than for controls (P .0003). No sig¬ =
nificant difference between the pa¬ tients with cerebellar ataxia and con¬ trols was seen for AChE activity. Phy¬ sostigmine failed to show any effect either on CSF AChE activity or CSF choline levels (Table 4). Levels of plasma amino compounds showed no significant alterations ei¬ ther due to disease (when compared with controls) or due to physostigmine (when compared with placebo). Cerebellar examinations videotaped during the 12th week of placebo and drug treatments were scored indepen¬ dently by three neurologists. The cor¬ relation coefficients among examiners were .85, .87, and .91. This was signif¬ icant (P < .001 level). For evaluating ataxia scores, the median score of three examiners for each patient dur¬ ing the placebo and drug phases of the study was compared between the
Lowering
of
scores
in¬
placebo and treated phases. The indi¬
vidual
scores
indicated that of six pa¬
tients, three showed improvement,
and three showed deterioration with nearly identical overall medians for the placebo phase (median = 25.5) and
drug phase (median 27.0); no signif¬ icant statistical change (P .5294) =
=
did functional im¬ provement occur in any of the patients was
observed
nor
(Figure). No significant side effects or changes
in the blood cell count and liver or re¬ nal functions attributable to the drug were noticed. There was no significant difference in the total protein, albu¬ min, glucose, sodium, potassium, chlo¬
ride, IgG, calcium,
phosphorus,
urea,
creatinine levels in the CSF ob¬ tained at the end of the study from either placebo- or physostigminetreated patients. or
COMMENT
The reduced CSF GABA level
re¬
ported in patients with cerebellar ataxia in this study confirms the re¬
sults of our earlier report1 and that of others.2 Furthermore, GABA levels in the CSF were lowest in patients with cerebellar ataxia compared with many other degenerative diseases, such as Huntington's disease, Parkinson's dis¬ ease, and Alzheimer's disease.1 In ad¬ dition to GABA, methionine levels in CSF were reduced in patients with
ethanolamine and reduced the levels of isoleucine. Studies have shown that following inhibition of GABA uptake by nipecotic acid in hippocampal slices, levels of phosphoethanolamine also increase.33 Similarly, inhibition of taurine uptake by guanidinoethane sulfonic acid in hippocampal slices re¬ sulted in an increase of extracellular taurine and phosphoethanolamine.34 Thus, it is possible that elevation of
phosphoethanolamine by physostig¬ mine may occur secondary to other neurochemical changes in the brain. Regarding the physostigmine-induced
reduction of CSF isoleucine levels, al¬ though the distribution of branch-
chain and other neutral amino acids among the blood, brain, and CSF may be affected by competing substances,35 no such interaction has been docu¬ mented with physostigmine. The selective reduction of GABA and methionine in CSF from patients with cerebellar ataxia suggests that specific chemical alterations that oc¬ cur in the central nervous system may be reflected in the CSF in this disease. Additionally, alterations of CSF phos¬ phoethanolamine and isoleucine observed following physostigmine treatment indicate the utility of CSF neurochemical measurements for monitoring effects of therapeutic strategies. Similar neurochemical al¬ terations occurring in CSF but not oc¬ curring in plasma have also been re¬ ported in patients with Huntington's disease who were treated with isoniazid.36 Neurochemical Studies
Neurochemical studies have shown that although AChE activity is partic¬ ularly high in the cerebellum of many species,3739 the activity of choline acetyltransferase39"13 and the levels of ACh4447 are low. Iontophoretic experi¬ ments suggest that synapses within the cerebellum of different species, not
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
only in the cortex but also in the deep cerebellar nuclei, may be cholinergic.48 Physostigmine, when administered to rats, inhibited cholinesterase activity uniformly in all brain areas, including the cerebellum; however, the levels of choline were not significantly altered.17
Thai et al49 administered physostig¬ mine salicylate orally to 12 patients with Alzheimer's disease in doses ranging between 0.5 and 2.5 mg every 2 hours, so that the daily dose never exceeded 16 mg. The CSF cholinest¬ erase inhibition measured 1 hour after the third daily dose showed a wide variation of values from 0% to 70% ; in our study, a maximum dose of 8 mg/d for 9 weeks (primed with 1 mg, 1 hour before a lumbar procedure) was ad¬ ministered to patients with cerebellar ataxia. The treated patients showed no change in CSF AChE activity. How¬ ever, the levels of choline in the CSF were significantly low. Several factors need to be considered. First, dietary intake of choline is negligible because none of our patients were on any dietary restriction. Second, it is un¬ likely that plasma choline levels had any influence on the data reported here, as it has been reported that plasma choline levels double during the daytime due to dietary intake.50 While we did not measure plasma choline levels, we performed lumbar puncture at the same time of the day (between 8 and 9 AM) in both the con¬ trol and experimental groups. Hence, the plasma choline level does not ex¬ plain the low CSF choline levels. We suggest that a decline in the CSF choline level may be due either to a de¬ fect in choline transport into the brain or to a decrease of choline-phospholipid output from the brain. The major source of brain choline is exogenous, and only a small amount is synthesized
by neuronal tissue.51 The origin of CSF choline is twofold: plasma choline and
choline from brain extracellular fluid. a reduced level of 'CSF choline may be due primarily to a lower uptake into the brain or secondarily to a lower release from the membrane choline-
Thus,
phospholipid pool. Alternatively, an increased active transport of choline out of the CSF into the blood at the level of the choroid plexus or at the blood-brain barrier could also explain
the decreased levels of this substance in the CSF. Choline, as a product of ACh hydrolysis, does not constitute a significant percent of the CSF choline pool. Further investigation into the specificity and mechanism of our find¬ ing should provide insight into choline metabolism and transport in patients with cerebellar ataxia. Physostigmine Treatment At the present time, there is no effective treatment of cerebellar atax¬ ias (for review, see Manyam52). The consistent finding of cholinergic defi¬ ciency in the brain of patients with in¬ herited ataxias53 and idiopathic cere¬ bellar degeneration has led to the at¬ tempt at a pharmacologie correction of cholinergic activity in the central ner¬ vous system. Choline, the precursor of ACh, has been used in the treatment of cerebellar ataxia, but the response has not always been favorable.54 58 Kark et al59 used low doses of physostigmine on patients with inherited ataxias and studied the effect in a double-blind, triple-crossover trial in 21 patients. It was reported that 13 (62%) showed consistent, statistically significant im¬ provement of ataxia scores by using
physostigmine compared with placebo. While the improvement was seen only in the ataxia scores, the patients did not experience subjective variation in
the ataxia during the trial. With the use of an identical dosage in the study reported here, the improvement in the score was minimal as evidenced only in three of six patients, with no change seen between the total scores of the placebo-treated group vs the physostigmine-treated group. Even the min¬ imal improvement of scores in three patients was not translated into func¬ tional improvement. While physostig¬ mine is an effective cholinergic agent and can be administred orally or parenterally, the risk of central and peripheral side effects from higher doses would limit its use. With the in¬ tent of minimizing or eliminating sys¬ temic side effects, physostigmine was administered intravenously after pre¬ treatment with systemic anticholin¬ ergic drugs, such as methscopolamine bromide, that do not cross the bloodbrain barrier, Even then, the side ef¬ fects were not fully preventable60 (B.V.M., unpublished data, 1977). In addition, the duration of action of oral physostigmine is probably about 2 hours or less. In view of this, it does not appear that higher doses of physostig¬ mine can be recommended or that physostigmine can be considered to be an effective treatment of cerebellar ataxia. This study was supported in part by grant AG05416 from the National Institute of Aging, by research support grant RR-5414 from the Na¬ tional Institutes of Health, by the Medical Re¬ search Program of the Veterans Affairs, Wash¬ ington, DC; and by the Percival E. and Ethyl Brown Foererder Foundation, Philadelphia, Pa. We wish to thank Elizabeth Williams for tech¬ nical assistance, Jerry A. Colliver, PhD, Division of Statistics and Research Consulting, Southern Illinois University School of Medicine, Spring¬ field, for statistical analysis, R. A. Pieter Kark, MD, for providing physostigmine and identical placebo tablets, and Leonard Katz, MD, and Don¬ ald Marcus, MD, for participation in blinded ataxia score evaluation.
References 1. Manyam NVB, Katz L, Hare TA, Gerber JC III, Grossman MH. Levels of \g=g\-aminobutyricacid in cerebrospinal fluid in various neurologic disorders. Arch Neurol. 1980;37:352-355. 2. Ogawa N, Kuroda H, Ota Z, Yammoto M, Otsuk S. Cerebrospinal fluid gamma-aminobutyric acid variations in cerebellar ataxia. Lancet. 1982;
2:215. 3. Perry TL, Currier RD, Hansen S, Maclean J. Aspartate-taurine imbalance in dominantly inherited olivopontocerebellar atrophy. Neurology.
7. Yamaguchi T, Hayashi K, Murakami H, Ota K, Maruyama S. Glutamate dehydrogenase deficiency in spinocerebellar degeneration. Neurochem Res. 1982;7:627-636. 8. Duvoisin R, Chokroverty S, Lepore E, Nicklas W. Glutamate dehydrogenase deficiency in patients with olivopontocerebellar atrophy. Neurology. 1983;33:1322-1326.
1977;27:257-261. 4. Perry TL, Kish SJ, Hansen S, Currier RD. Neurotransmitter amino acids in dominantly inherited disorders. Neurology. 1981;31:237-242. 5. Perry TL, Kish SJ, Hinton D, Hansen S, Becker LE, Gelfand EW. Neurochemical abnormalities in a patient with ataxia-telangiectasia. Neurology. 1984;34:187-191. 6. Plaitakis A, Nicklas WJ, Desnick RJ. Glutamate dehydrogenase deficiency in three patients with spinocerebellar syndrome. Ann Neurol. 1980;
9. Obata K. Association of GABA with cerebellar Purkinje cells: a single cell analysis. In: Roberts E, Chase T, Tower D, eds. GABA in Nervous System Function. New York, NY: Raven Press; 1980:217-222. 10. Storm-Mathisen J. Distribution of the components of the GABA system in neuronal tissue: cerebellum and hippocampus\p=m-\effectsof axotomy. In: Roberts E, Chase T, Tower D, eds. GABA in Nervous System Function. New York, NY: Raven Press; 1976:149-168. 11. Nadi NS, McBride WJ, Aprison MH. Distribution of several amino acids in regions of the cerebellum of the rat. J Neurochem. 1977;28:453\x=req-\
7:297-303.
455.
12. Roffler-Tarlov S, Sidman RL. Concentrations of glutamic acid in cerebellar cortex and deep nuclei of normal mice and weaver, stagger, and nervous mutants. Brain Res. 1978;142:269\x=req-\ 283. 13. Nadi NS, Kanter D, McBride WJ. Effects of 3-acetylpyridine on several putative neurotransmitter amino acids in the cerebellum and medulla of the rat. J Neurochem. 1977;28:661-662. 14. Kish SJ, Perry TL, Hornykiewicz O. Increased GABA receptor binding in dominantly inherited cerebellar ataxias. Brain Res. 1983; 269:370-373. 15. Blass JP, Gibson GE: Studies in pathophysiology of pyruvate dehydrogenase deficiency. In: Kark RAP, Rosenberg RN, Schut LJ, eds. The Inherited Ataxias. New York, NY: Raven Press; 1978:181-194. 16. Kish SJ, Currier RD, Schut L, Perry TL, Morito CL. Brain choline acetyltransferase reduction in dominantly inherited olivopontocerebellar atrophy. Ann Neurol. 1987;22:272-275. 17. Hallak M, Giacobini E. Relation of brain
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
regional physostigmine concentration to cholinesterase activity and acetylcholine and choline
levels in rat. Neurochem Res. 1986;2:1037-1048. 18. Kark RA, Blass JP, Spence MA. Physostigmine in familial ataxias. Neurology. 1977;27:70\x=req-\ 72. 19. Manyam NVB, Hare TA, Katz L. Effect of isoniazid on cerebrospinal fluid and plasma GABA levels in Huntington's disease. Life Sci.
1980;26:1303-1308.
20. Deutsch SI, Mohs RC, Levy MI, et al. Acetylcholinesterase activity in CSF in schizophrenia, depression, Alzheimer's disease, and normals. Biol Psychiatry. 1983;18:1363-1373. 21. Grossman MH, Hare TA, Manyam NVB, Glaeser BS, Wood JH. Stability of GABA levels in
CSF under various conditions of storage. Brain Res. 1980;182:99-106. 22. Lundquist C, Blomstrand C, Hamberger A, Wikkelso C. Liquid chromatographic separation of cerebrospinal fluid amino acids after precolumn fluorescence derivatization. Acta Neurol Scand. 1989;79:273-279. 23. Ferraro TN, Hare TA. Triple-column ionexchange physiological amino acid analysis with fluorescent detection: baseline characterization of human CSF. Anal Biochem. 1984;143:82-94. 24. Johnson CD, Russell RL. A rapid, simple radiometric assay for cholinesterase. Anal Biochem. 1975;64:229-238. 25. Elble R, Giacobini E, Scarsella GF. Cholinesterases in cerebrospinal fluid: a longitudinal study in Alzheimer disease. Arch Neurol.
1987;44:403-407.
26. McCaman RE, Stetzler J. Radiochemical acetylcholine: modification for subpicomoles measurements. J Neurochem. 1977; 28:669-671. 27. SAS User's Guide: Statistics. Cary, NC: SAS Institute; 1982. 28. Perry TL, Berry K, Hansen S, Diamond S, Mok C. Regional distribution of amino acids in human brain obtained at autopsy. J Neurochem. assay for
1971;18:513-519. 29. Perry TL, Sanders HD, Hansen S, Lesk D, Kloster M, Graslin L. Free amino acids and related compounds in five regions of biopsied cat brain. J Neurochem. 1972;19:2651-2656. 30. Shaw RK, Heine JD. Ninhydrin positive
substances present in different areas of normal rat brain. J Neurochem. 1965;12:151-155. 31. Tallan HH, Rassin DK, Starman JA, et al.
Methionine metabolism in the brain. In: Lajtha A1 ed. Handbook of Neurochemistry. 2nd ed. New York, NY: Plenum Press; 1983;3:535-538. 32. Cohen SM, Nichols A, Wyatt R, Pullim W. The administration of methionine in chronic schizophrenic patients: a review of ten studies. Br J Psych. 1974;8:209-225. 33. Hamberger A, Berthold A, Jacobson I, et al. In vivo brain dialysis of extracellular nontransmitters on putative transmitter amino acids. In: Bayon A, Drucker-Colin R, eds. In Vivo Perfusion and Release of Neuroactive Substances. Orlando, Fla: Academic Press Inc; 1985:119-139. 34. Lehman A, Hamberger A. A possible interrelationship between extracellular taurine and phosphoethanolamine in the hippocampus. J Neurochem. 1984;42:1286-1290. 35. Oldendorf WH. Saturation of blood brain barrier transport of amino acids in phenylketonuria. Arch Neurol. 1973;28:45-48. 36. Manyam BV, Katz L, Hare TA, Kaniefski K, Tremblay RD. Isoniazid induced elevation of CSF GABA levels and effects on chorea in Huntington's disease. Ann Neurol. 1981;10:35-37. 37. Burgen ASV, Chipman LM. Cholinesterase and succinic dehydrogenase in the central nervous system of the dog. J Physiol. 1951;114:296-305. 38. Austin L, Phillis JW. The distribution of cerebellar cholinesterase in several species. J Neurochem. 1965;12:709-717. 39. Kasa P, Silver A. The correlation between choline acetyltransferase and acetylcholinesterase activity in different areas of the cerebellum of rat and guinea pig. J Neurochem. 1969;16:389\x=req-\ 396. 40. Feldberg W, Vogt M. Acetylcholine synthesis in different regions of the central nervous system. J Physiol. 1948;107:372-381. 41. Hebb CO, Silver A. Choline acetylase in the central nervous system of man and some other mammals. J Physiol. 1956;134:718-728. 42. Goldberg AM, McCaman RE. A quantitative microchemical study of choline acetyltransferase and acetylcholinesterase in the cerebellum of several species. Life Sci. 1967;6:1493-1500. 43. Kasa P. The cholinergic systems in brain and spinal cord. Prog Neurobiol. 1986;26:211-272. 44. MacIntosh FC. The distribution of acetylcholine in the peripheral and the central nervous system. J Physiol (Lond). 1941;99:436-442. 45. Welsh JH, Hyde JE. The distribution of acetylcholine in brains of rats of different ages.
J Neurophysiol. 1944;7:41-49. 46. Vizi SE, Palkovits M. Acetylcholine content in different regions of the rat brain. Brain Res Bull. 1978;3:93-96. 47. Kasa P, Banasghy K, Rakonczay Z, Gulya K. Postnatal development of the acetylcholine system in different parts of the rat cerebellum. J Neurochem. 1982;39:1726-1732. 48. Kasa P. The cholinergic systems in brain and spinal cord. Prog Neurobiol. 1986;26:211-272. 49. Thal LJ, Masur DM, Fuld PA, Sharpless NS, Davies P. Memory improvement with oral physostigmine and lecithin in Alzheimer's disease. In: Katzman R, ed. Banbury Report 15: Biological Aspects of Alzheimer's Disease. Cold Spring Harbor, Mass: Cold Spring Laboratory; 1983:461-469. 50. Zeisel SH, Growdon JH, Wurtman RJ, Magil SG, Logue M. Normal plasma choline responses to
ingested lecithin. Neurology. 1980;
30:1226-1229. 51. Blusztajn JK, Wurtman RJ. Choline and cholinergic neurons. Science. 1983;221:614-620. 52. Manyam BV. Recent advances in the treatment of cerebellar ataxias. Clin Neuropharmacol.
1986;9:508-516.
53. Barbeau A. Emerging treatments: replacetherapy with choline or lecithin in neurological diseases. Can J Neurol Sci. 1978;5:157-160. 54. Lawrence CM, Millac P, Stout GS, Ward JW. The use of choline chloride in ataxic disorders. J Neurol Neurosurg Psychiatry. 1980;43:452\x=req-\ 454. 55. Legg NJ. Oral choline in cerebellar ataxia. BMJ. 1978;2:1403-1404. 56. Legg N. Oral choline in cerebellar ataxia. BMJ. 1979;2:133. 57. Livingstone IR, Mastaglia FL, Pennington RJT, Skilbeck C. Choline chloride in the treatment of cerebellar and spinocerebellar ataxia. J Neurol Sci. 1981;50:161-174. 58. Schested P, Lund HI, Kristensen 0. Oral choline in cerebellar ataxia. Acta Neurol Scand. ment
1980;62:124-126. 59. Kark RA, Budelli MMR, Wachsner R. Double-blind triple-cross trial of low doses of oral physostigmine in inherited ataxia. Neurology. 1981;31:288-292. 60. Budelli MMR, Kark RAP, Blass JP, Spence
MA. Action of physostigmine of inherited ataxias. Adv Neurol. 1978;21:195-202.
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
Giacobini, PhD; Thomas N. Ferraro, PhD; Theodore A. Hare, PhD
\s=b\ Cerebrospinal fluid (CSF) amino acid neurotransmitters, related compounds, and their precursors, choline levels, and acetycholinesterase activity were measured in the CSF of patients with cerebellar ataxia during a randomized, double\x=req-\ blind, crossover, placebo-controlled clinical trial of physostigmine salicylate. The CSF \g=g\-aminobutyricacid, methionine, and choline levels, adjusted for age, were significantly lower in patients with cerebellar ataxia compared with controls. Physostig-
mine selectively reduced the level of CSF isoleucine and elevated the levels of phosphoethanolamine. No change occurred in CSF acetylcholinesterase activity and in the levels of plasma amino compounds in patients with cerebellar ataxia when compared with controls. Median ataxia scores did not statistically differ between placebo and physostigmine nor did functional improvement occur in any of the patients. (Arch Neurol. 1990;47:1194-1199)
í^ erebellar ataxia represents
a diffuse group of disorders for which both inherited and noninherited forms have been described. Age at onset is variable but usually occurs in middle age. Studies on the cerebrospinal fluid (CSF) neurochemistry of noninherited
Accepted for publication March 16, 1990. From the Division of Neurology (Dr Manyam) and Department of Pharmacology (Dr Giacobini), Southern Illinois University School of Medicine, Springfield, and the Departments of Pharmacology (Drs Ferraro and Hare) and Neurology (Dr Ferraro), Thomas Jefferson University, Philadelphia, Pa. Reprint requests to Division of Neurology, D411, Southern Illinois University School of Medicine, PO Box 19230, Springfield, IL 62794-9230 (Dr Manyam).
cerebellar disorders
are
scarce1-2; how¬
ever, abnormalities of amino acid
neu¬
rotransmitter systems have been
ported in postmortem brains of
re¬
pa¬
tients who died with inherited forms of cerebellar dysfunction.35 Deficiency of glutamate dehydroge¬ nase has been reported in a subset of patients with olivopontocerebellar atrophy.68 Glutamate dehydrogenase forms part of a complex enzymatic process for transferring amino groups between a-keto acids and their corre¬ sponding -amino acids. Furthermore, investigation of amino acids in cere¬ bellar disorders is particularly appro¬ priate since at least seven major types of cerebellar neurons are believed to use amino acids as their neuro¬ transmitters.913 -Aminobutyric acid (GABA) is believed to be the inhibitory neurotransmitter of Purkinje's cells9 and may also be used by the three in¬ hibitory interneurons in the cerebellar cortex—the basket and the stellate and Golgi's cells10—although it has been suggested that taurine rather than GABA may be the inhibitory neurotransmitter used by the stellate cells.11 Glutamate is believed to be the neurotransmitter used by the mossy and climbing fibers and by the granule cells with their parallel fibers,12 al¬ though evidence also exists to impli¬ cate aspartate as the neurotransmit¬ ter of parallel fibers.13 A study of amino acids in postmor¬ tem brain tissue from 10 patients (six different pedigrees) with dominantly inherited cerebellar disorders" re¬ vealed three biochemically different disorders despite clinical similarities. One disorder was characterized by
moderate reduction of aspartate and
glutamate levels in cerebellar cortex; a second disorder was characterized by reductions of aspartate and glutamate levels in other brain regions in addi¬
tion to cerebellar cortex; and a third disorder was characterized by normal aspartate and glutamate concentra¬ tions in all brain regions examined. The GABA content in cerebellar cortex and dentate nucleus was reduced in some patients with each disorder, whereas cerebellar taurine was nor¬ mal in all patients. Increased GABA receptor binding in the cerebellar cor¬ tex of patients who died with cerebel¬ lar ataxias has been documented"1 and this finding is consistent with de¬ creased GABA levels in the dentate nucleus.4 While the full CSF amino acid profile for cerebellar ataxia has not been previously reported, it has been documented that CSF GABA is significantly reduced in this disorder.1·2 Several inherited ataxias are asso¬ ciated with defects of pyruvate oxida¬ tion, and such defects are known to re¬ sult in the inhibition of acetylcholine (ACh) synthesis in the brain.15 The ac¬ tivity of choline acetyltransferase, the enzyme that synthesizes ACh from choline, and ACh, are markedly re¬ duced in the brain of patients who die with olivopontocerebellar atrophy.16 Physostigmine salicylate has a revers¬ ible inhibitory effect on cholinesterase; as a consequence, it enhances cholin¬ ergic transmission at synapses. The drug is readily absorbed from the gas¬ trointestinal tract, easily crosses the blood-brain barrier, and acts on the central nervous system." In a doubleblind, randomized, clinical trial of
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
physostigmine, we evaluated possible alterations of cholinergic function and
Table 1.—Physostigmine Treatment of Cerebellar Ataxia: Clinical
amino acid neurotransmitters in pa¬ tients with cerebellar ataxia by mea¬ suring levels of CSF amino acid neuro¬ transmitters, related compounds, their precursors, and acetylcholinesterase
Patient
Summary Age, y Family history
(AChE) activity.
PATIENTS AND METHODS
seen on a
±
SD age of 53
±
cholinergic agonistic or antagonistic drugs. No dietary restrictions were im¬ posed. Informed consent was obtained from patients, and approval for the trial was granted by the Institutional Committee on Human Experimentation. To study the effect of physostigmine on any
CSF neurochemicals and cerebellar symp¬ toms, physostigmine was administered with the use of a placebo-control, doubleblind crossover design so that the CSF col¬ lected during the placebo phase would re¬ flect changes due to disease, and any addi¬ tional changes in CSF collected during the physostigmine phase would indicate changes due to treatment. Physostigmine salicylate (1 mg) or identical placebo (lac¬ tose) tablets were allotted to patients with ataxia who were chosen randomly for 12 weeks each. At the end of 12 weeks, physo¬ stigmine and placebo were crossed over for another 12 weeks. Either physostigmine or placebo was administered at a rate of one tablet three times each day during the first week, one tablet four times each day during the second week, one tablet six times each day during the third week, and then one tablet given once every 2 hours, except while asleep (eight times each day) from the fourth through the 12th week. Before the start of the trial, patients were videotaped and then retaped during the 12th week of both placebo and drug treatment. All video tape recording was done at approximately the same time of day. One hour before recording time, a dose of either physostig¬ mine (or placebo) was administered so as to induce a maximum therapeutic effect. Cerebellar function was assessed by semiquantitative scoring as follows: 0, absence of ataxia; 1, minimal or extreme normal; 2, mild; 3, moderate; 4, severe; and 5, inability to carry out the test.18 The total score for each observation was used for comparison. Reduction in score indicated improvement. Complete blood cell counts, liver and renal function test results, serum electrolyte de¬ terminations, and calcium and phosphorus
62
47
52
35
53 ± 11 + 0/6
Alcohol intake Skeletal deformities
0/6
Dementia
0/6 5/6
83
4/6 6/6 6/6
67 100 100
6/6
100
Intention tremor
Dysdladochoklnesia Truncal ataxia Gatt ataxia
computed tomographic (CT) scan.
Patients 2, 3, 4, and 6 gave a history of oc¬ casional alcohol intake, but other features of chronic alcoholism, such as dementia, peripheral neuropathy, and reversal of al¬ bumin-globulin ratios in the serum, were lacking. None of the patients were receiving
62
Optic atrophy Dysarthria Nystagmus Dysmetria, F-N Dysmetria, H-S
11 years (range, 35 to 62 years) participated in the trial. The details of clinical data are presented in Table 1. All patients were con¬ sidered to have sporadic cerebellar ataxia, except patient 1, in whom early olivoponto¬ cerebellar atrophy could not be excluded as an extensor plantar response with cerebel¬ lar atrophy without brain-stem atrophy as
Six men with a mean
Findings*
83
Muscle tone DTRs Plantar responses CT scan, cerebellar
i
1
Hyper
Hypo
Hypo
tt
il
li
atrophy
u li
ii
ii
17 4/6
0
67
*
Asterisk indicates unable to test; 0, negative or none; plus sign, present or positive; F-N, finger to nose test; H-S, heel to shin test; N, within normal limits; DTRs, deep tendon reflexes; Hyper, hyperactive; Hypo, hypoactlve; upward arrow, extensor response; downward arrow, flexor response; an arrow, each lower limb; and CT,
computed tomographic. tMean ± SD.
levels riod.
were
monitored
during the trial
pe¬
Lumbar Puncture
Cerebrospinal fluid was collected by lum¬ bar puncture at the end of 12 weeks of both placebo and physostigmine treatment phases. The CSF was collected between 8 and 9 am in all patients following overnight bed rest and fasting (water was allowed ad libitum to prevent dehydration) for 12 hours. In patients under treatment for cerebellar ataxia, either placebo or physo¬ stigmine salicylate (1 mg) was adminis¬ tered 1 hour before CSF collection. Lumbar puncture was performed in the standard manner. The details of CSF collection, prep¬ aration, and storage conditions have been described elsewhere.19 Cerebrospinal fluid for measurement of neurochemicals from both controls and the experimental group was from the same gradient (first 10 mL of CSF) and, thus, was not subject to gradient
Acetylcholinesterase retains enzyme activity in CSF after long-term storage at -70°C and even after repeated freezing and thawing.20 The reliability of freezing naive CSF before high-performance liquid chro¬ matography analysis for CSF amino com¬ pounds and of using the same gradient from effect.
the first 10 mL of CSF has been confirmed in our laboratory21 and others.22 Control CSF was obtained from 17 men without evidence of organic neurologic dis¬ ease or mental disorders. Of these, CSF samples from 10 men with a mean ± SD age of 48 ± 20 years were used to act as controls for amino compound studies, and samples from nine men with a mean ±SD age of 64 ± 9 years were controls for choline lev¬ els and AChE activity studies with an over¬ lap of two controls. This was done due to the limited availability of CSF from controls,
as
the CSF from the first 10 mL was used for
comparison between the controls and ex¬ perimental group to control gradient effect. Protein, glucose, sodium, potassium, chloride, calcium, and phosphorus values were measured in the clinical laboratory with an automated analyzer. The IgG level was
measured with radioimmunodiffusion. Amino
Compound Analysis
Cerebrospinal fluid and plasma amino were quantified with the use of a tri¬ ple-column, ion-exchange Chromatographie procedure with fluorometric detection.23 Cerebrospinal fluid and plasma proteins were precipitated by adding a one-third volume of 20% aqueous sulfosalicylic acid. Specimens were centrifuged at 2000¡7, and acids
the supernatant fluids were used for direct injection onto the high-performance liquid Chromatograph. Norleucine and 5-aminovaleric acid were used as internal stan¬ dards. The analytical procedure used par¬ allel operation of three distinct ionexchange columns for separation and mea¬ surement of acidic, neutral, and basic amino acids, respectively, with a recovery rate for an internal standard of 98% and a lower limit of detection of 5 pmol. The chromatography was run isocratically in each case with the use of lithium citrate buffers designed to provide overlapping elution profiles. All samples were run in duplicate, and the mean was used for calcu¬ lation.
Assay of AChE Activity The radiometrie method of Johnson and Russell24 was used to determine AChE ac¬ tivity with tritiated ACh iodide as a sub¬
(90 mCi/mmol, specific activity). Acetylcholinesterase activity was exstrate
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
Table 2.—CSF Amino
Compounds
pressed as micromoles of tritiated ACh io¬ dide hydrolyzed per hour per milliliter of
in Cerebellar Ataxia* Mean ±
SD,
nmol/mL
Patients With Cerebellar Ataxia ( 6) 0.051 ± 0.024
Controls Amino Compounds /3-Alanlne
0.078 ± 0.035
GABA
0.142 ± 0.035
Aspartate 1-Methylhlstidine 3-Methylhlstldine
0.275 ± 0.151
0.084 ± 0.012t 0.223 ± 0.159
1.24 ± 1.12
1.19 ± 0.736 1.20 ± 0.484
(n
=
10)
=
Homocarnosine
1.30 ± 1.01 1.84 ± 1.10
Tryptophan
2.75 ± 1.67
0.855 ± 0.421 2.76 ± 1.25
Methyl-lysine
2.83 ± 3.10
3.44 ± 2.62
Citrulline
2.88 ± 1.02
2.63 ± 0.591
tj-Amlnobutyrate
3.31
3.90 ± 1.34
1.08
Glutamate
3.13 ± 0.395 3.51 ± 0.729 3.76 ± 0.957
Phosphoethanolamine Methionlne
3.09 ± 0.609
2.37 ± 0.517t
Ornithine Isoleuclne Taurine
4.63 ± 0.664
5.29 ± 0.915
4.95 ± 1.18
5.99 ± 0.740 5.17 ± 1.75
Glycine
8.08 ± 4.26
7.15 ± 0.631
Asparagine
9.24 ± 1.93
9.08 ± 1.49 8.87 ± 1.48
Ethanolamlne
9.01
Phenylalanine Tyrosine
10.16 ± 2.62 10.73 ± 5.19
8.32 ± 2.36
Histidlne
12.06 ± 2.53 13.28 ± 2.76
12.13 ± 2.73 13.75 ± 1.87
Valine
16.69 ± 4.01
19.88 ± 2.58
Arginine Serine
22.25 ± 4.01 25.41 ± 4.79
25.65 ± 4.01
Lysine
25.69 ± 4.46
29.37 ± 2.74
Alanine Threonine Glutamine
30.44 ± 6.90 33.77 ± 4.18
Leuclne
9.58 ± 2.43
29.27 ± 4.81
34.48 ± 5.06 655.3 ± 63.6
311.5 ± 67.2
*CSF indicates cerebrospinal fluid; GABA, -aminobutyric acid. tBy analysis of covariance with age as covariate (P .0001).
Table 3.—Effect of
=
SD, nmol/mL
Phosphoethanolamine
Isoleucine
3.09 ± 0.61 3.64 ± 0.67
5.99 ± 0.74
.0072
.0209
Physostigmine salicylate Slgnlflcancet *CSF indicates cerebrospinal fluid. tBy paired Student's i test.
Controls Patients with cerebellar ataxia
Significance, P(CSF AChE)
0.28
0.75 ± 0.20 .2440t Physostigmine sallcylatet 6 0.74 ± 0.21 ,5678§ *CSF Indicates cerebrospinal fluid; AChE, acetylcholinesterase. tBy analysis of covariance with age as covariate. tDouble-blind crossover trial with final dose at 8 mg/d for 9 weeks. §By paired Student's t test. Placebo
•
ease
(1.36/0.03 mmol-mL-h).25 Choline Assay
For the assay of CSF choline, 1 mL of CSF mixed with 10 mL of IN formic acidacetone (ratio, 15:85) and centrifuged at 30000 for 20 minutes to remove protein. The levels of choline were measured in 15-uL aliquots of CSF by the radiometrie enzy¬ matic assay.26 Absolute levels of choline were calculated on the basis of standards (10,50, and 100 pmol). All samples were as¬ sayed in duplicate. was
Statistical Analysis
Computations were done on an IBM 4341
ance with age as covariate was used to con¬ trol the effect of age on CSF and plasma components between controls and patients with cerebellar ataxia. The paired t test was applied to compare the effect of physostig¬ mine between placebo and treated phases, as this was a crossover trial. On ataxia rat¬ ing scores, the Wilcoxon matched-pairs signed-rank test was used.
RESULTS
Table 4.—Cerebellar Ataxias: CSF AChE Activity Choline Levels* CSF AChE, mL µ per h
of 10_i mol/L. At this concentration, isoompa inhibited 90% BuChE CSF activity measured with tritiated butyrylcholine io¬ dide as a substrate. In CSF, BuChE activity represents approximately 5% of total cholinesterase activity, and there is no dif¬ ference, for example, in the ratio of AChE:BuChE for controls (1.35/0.061 mmol mL-h) vs patients with Alzheimer's dis¬
procedure27 was used to estimate means, SDs, and significance. Analysis of covari¬
on CSF Amino Compounds in Patients With Cerebellar Ataxias ( 6)*
Placebo
tetraisopropylpyrophosphoramide, Sigma Chemical Co, St Louis, Mo) of 10"3 to 10"" mol/L. Based on these experimental re¬ sults, to inhibit BuChE activity present in CSF, iso-ompa was used at a concentration
computer. The Statistical Analysis System
Physostigmine
Mean ±
CSF. To determine the proper concentra¬ tion of the inhibitor, inhibition experi¬ ments were performed on CSF, as well as on purified samples of butyrylcholinesterase (BuChE, serum) and AChE (erythrocytes), at decreasing concentration of inhibitors (BW284C51, Burroughs Wellcome Co, Re¬ search Triangle Park, NC, and iso-ompa,
CSF
Choline, nmol/mL
Significance, /»(CSF Choline)
2.97 ± 0.79
1.02 ± 0.29 0.96 ± 0.19§
.0003t 3825§
To study the alteration of CSF amino compounds in patients with cerebellar ataxia, CSF obtained during the placebo phase was compared with that of controls. During the placebo phase, patients were monitored closely for 12 weeks before CSF collection, and thus, the values could be considered to
represent
a more
homogeneous
popu¬
lation than that represented by ran¬ domly collected samples. To study the effect of physostigmine on CSF and plasma amino compounds, values gen¬ erated from samples collected at the end of 12 weeks of placebo and 12 weeks of physostigmine were compared. With the use of analysis of covariance, with age as covariate, a significant re-
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
cerebellar ataxia. Methionine derived from diet or other tissues crosses the blood-brain barrier, and differences in methionine levels in the various brain regions are not great.28 30 Methionine in the free amino acid pool of the brain appears to be used mainly for protein synthesis.31 Methionine has been shown to have a significant effect when administered to schizophrenic pa¬ tients, causing a marked aggravation of symptoms.32 While this establishes the importance of methionine in cen¬ tral nervous system function, its role in cerebellar function needs further
study. Physostigmine treatment selec¬ tively elevated the levels of phospho-
Effect of oral physostigmine treatment (dotted bars) on cerebellar ataxia. dicates improvement; open bars, placebo treatment.
duction ( .0001) of CSF GABA and methionine was found in patients with cerebellar ataxia when compared with controls (Table 2). Physostigmine se¬ lectively elevated the levels of phosphoethanolamine (P .0072) and re¬ duced the levels of isoleucine (P .0209) but had no significant ef¬ fect on levels of other CSF amino com¬ =
=
=
pounds (Table 3).
Mean ± SD CSF choline levels, ad¬
justed for age, were significantly lower for patients with cerebellar ataxia than for controls (P .0003). No sig¬ =
nificant difference between the pa¬ tients with cerebellar ataxia and con¬ trols was seen for AChE activity. Phy¬ sostigmine failed to show any effect either on CSF AChE activity or CSF choline levels (Table 4). Levels of plasma amino compounds showed no significant alterations ei¬ ther due to disease (when compared with controls) or due to physostigmine (when compared with placebo). Cerebellar examinations videotaped during the 12th week of placebo and drug treatments were scored indepen¬ dently by three neurologists. The cor¬ relation coefficients among examiners were .85, .87, and .91. This was signif¬ icant (P < .001 level). For evaluating ataxia scores, the median score of three examiners for each patient dur¬ ing the placebo and drug phases of the study was compared between the
Lowering
of
scores
in¬
placebo and treated phases. The indi¬
vidual
scores
indicated that of six pa¬
tients, three showed improvement,
and three showed deterioration with nearly identical overall medians for the placebo phase (median = 25.5) and
drug phase (median 27.0); no signif¬ icant statistical change (P .5294) =
=
did functional im¬ provement occur in any of the patients was
observed
nor
(Figure). No significant side effects or changes
in the blood cell count and liver or re¬ nal functions attributable to the drug were noticed. There was no significant difference in the total protein, albu¬ min, glucose, sodium, potassium, chlo¬
ride, IgG, calcium,
phosphorus,
urea,
creatinine levels in the CSF ob¬ tained at the end of the study from either placebo- or physostigminetreated patients. or
COMMENT
The reduced CSF GABA level
re¬
ported in patients with cerebellar ataxia in this study confirms the re¬
sults of our earlier report1 and that of others.2 Furthermore, GABA levels in the CSF were lowest in patients with cerebellar ataxia compared with many other degenerative diseases, such as Huntington's disease, Parkinson's dis¬ ease, and Alzheimer's disease.1 In ad¬ dition to GABA, methionine levels in CSF were reduced in patients with
ethanolamine and reduced the levels of isoleucine. Studies have shown that following inhibition of GABA uptake by nipecotic acid in hippocampal slices, levels of phosphoethanolamine also increase.33 Similarly, inhibition of taurine uptake by guanidinoethane sulfonic acid in hippocampal slices re¬ sulted in an increase of extracellular taurine and phosphoethanolamine.34 Thus, it is possible that elevation of
phosphoethanolamine by physostig¬ mine may occur secondary to other neurochemical changes in the brain. Regarding the physostigmine-induced
reduction of CSF isoleucine levels, al¬ though the distribution of branch-
chain and other neutral amino acids among the blood, brain, and CSF may be affected by competing substances,35 no such interaction has been docu¬ mented with physostigmine. The selective reduction of GABA and methionine in CSF from patients with cerebellar ataxia suggests that specific chemical alterations that oc¬ cur in the central nervous system may be reflected in the CSF in this disease. Additionally, alterations of CSF phos¬ phoethanolamine and isoleucine observed following physostigmine treatment indicate the utility of CSF neurochemical measurements for monitoring effects of therapeutic strategies. Similar neurochemical al¬ terations occurring in CSF but not oc¬ curring in plasma have also been re¬ ported in patients with Huntington's disease who were treated with isoniazid.36 Neurochemical Studies
Neurochemical studies have shown that although AChE activity is partic¬ ularly high in the cerebellum of many species,3739 the activity of choline acetyltransferase39"13 and the levels of ACh4447 are low. Iontophoretic experi¬ ments suggest that synapses within the cerebellum of different species, not
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
only in the cortex but also in the deep cerebellar nuclei, may be cholinergic.48 Physostigmine, when administered to rats, inhibited cholinesterase activity uniformly in all brain areas, including the cerebellum; however, the levels of choline were not significantly altered.17
Thai et al49 administered physostig¬ mine salicylate orally to 12 patients with Alzheimer's disease in doses ranging between 0.5 and 2.5 mg every 2 hours, so that the daily dose never exceeded 16 mg. The CSF cholinest¬ erase inhibition measured 1 hour after the third daily dose showed a wide variation of values from 0% to 70% ; in our study, a maximum dose of 8 mg/d for 9 weeks (primed with 1 mg, 1 hour before a lumbar procedure) was ad¬ ministered to patients with cerebellar ataxia. The treated patients showed no change in CSF AChE activity. How¬ ever, the levels of choline in the CSF were significantly low. Several factors need to be considered. First, dietary intake of choline is negligible because none of our patients were on any dietary restriction. Second, it is un¬ likely that plasma choline levels had any influence on the data reported here, as it has been reported that plasma choline levels double during the daytime due to dietary intake.50 While we did not measure plasma choline levels, we performed lumbar puncture at the same time of the day (between 8 and 9 AM) in both the con¬ trol and experimental groups. Hence, the plasma choline level does not ex¬ plain the low CSF choline levels. We suggest that a decline in the CSF choline level may be due either to a de¬ fect in choline transport into the brain or to a decrease of choline-phospholipid output from the brain. The major source of brain choline is exogenous, and only a small amount is synthesized
by neuronal tissue.51 The origin of CSF choline is twofold: plasma choline and
choline from brain extracellular fluid. a reduced level of 'CSF choline may be due primarily to a lower uptake into the brain or secondarily to a lower release from the membrane choline-
Thus,
phospholipid pool. Alternatively, an increased active transport of choline out of the CSF into the blood at the level of the choroid plexus or at the blood-brain barrier could also explain
the decreased levels of this substance in the CSF. Choline, as a product of ACh hydrolysis, does not constitute a significant percent of the CSF choline pool. Further investigation into the specificity and mechanism of our find¬ ing should provide insight into choline metabolism and transport in patients with cerebellar ataxia. Physostigmine Treatment At the present time, there is no effective treatment of cerebellar atax¬ ias (for review, see Manyam52). The consistent finding of cholinergic defi¬ ciency in the brain of patients with in¬ herited ataxias53 and idiopathic cere¬ bellar degeneration has led to the at¬ tempt at a pharmacologie correction of cholinergic activity in the central ner¬ vous system. Choline, the precursor of ACh, has been used in the treatment of cerebellar ataxia, but the response has not always been favorable.54 58 Kark et al59 used low doses of physostigmine on patients with inherited ataxias and studied the effect in a double-blind, triple-crossover trial in 21 patients. It was reported that 13 (62%) showed consistent, statistically significant im¬ provement of ataxia scores by using
physostigmine compared with placebo. While the improvement was seen only in the ataxia scores, the patients did not experience subjective variation in
the ataxia during the trial. With the use of an identical dosage in the study reported here, the improvement in the score was minimal as evidenced only in three of six patients, with no change seen between the total scores of the placebo-treated group vs the physostigmine-treated group. Even the min¬ imal improvement of scores in three patients was not translated into func¬ tional improvement. While physostig¬ mine is an effective cholinergic agent and can be administred orally or parenterally, the risk of central and peripheral side effects from higher doses would limit its use. With the in¬ tent of minimizing or eliminating sys¬ temic side effects, physostigmine was administered intravenously after pre¬ treatment with systemic anticholin¬ ergic drugs, such as methscopolamine bromide, that do not cross the bloodbrain barrier, Even then, the side ef¬ fects were not fully preventable60 (B.V.M., unpublished data, 1977). In addition, the duration of action of oral physostigmine is probably about 2 hours or less. In view of this, it does not appear that higher doses of physostig¬ mine can be recommended or that physostigmine can be considered to be an effective treatment of cerebellar ataxia. This study was supported in part by grant AG05416 from the National Institute of Aging, by research support grant RR-5414 from the Na¬ tional Institutes of Health, by the Medical Re¬ search Program of the Veterans Affairs, Wash¬ ington, DC; and by the Percival E. and Ethyl Brown Foererder Foundation, Philadelphia, Pa. We wish to thank Elizabeth Williams for tech¬ nical assistance, Jerry A. Colliver, PhD, Division of Statistics and Research Consulting, Southern Illinois University School of Medicine, Spring¬ field, for statistical analysis, R. A. Pieter Kark, MD, for providing physostigmine and identical placebo tablets, and Leonard Katz, MD, and Don¬ ald Marcus, MD, for participation in blinded ataxia score evaluation.
References 1. Manyam NVB, Katz L, Hare TA, Gerber JC III, Grossman MH. Levels of \g=g\-aminobutyricacid in cerebrospinal fluid in various neurologic disorders. Arch Neurol. 1980;37:352-355. 2. Ogawa N, Kuroda H, Ota Z, Yammoto M, Otsuk S. Cerebrospinal fluid gamma-aminobutyric acid variations in cerebellar ataxia. Lancet. 1982;
2:215. 3. Perry TL, Currier RD, Hansen S, Maclean J. Aspartate-taurine imbalance in dominantly inherited olivopontocerebellar atrophy. Neurology.
7. Yamaguchi T, Hayashi K, Murakami H, Ota K, Maruyama S. Glutamate dehydrogenase deficiency in spinocerebellar degeneration. Neurochem Res. 1982;7:627-636. 8. Duvoisin R, Chokroverty S, Lepore E, Nicklas W. Glutamate dehydrogenase deficiency in patients with olivopontocerebellar atrophy. Neurology. 1983;33:1322-1326.
1977;27:257-261. 4. Perry TL, Kish SJ, Hansen S, Currier RD. Neurotransmitter amino acids in dominantly inherited disorders. Neurology. 1981;31:237-242. 5. Perry TL, Kish SJ, Hinton D, Hansen S, Becker LE, Gelfand EW. Neurochemical abnormalities in a patient with ataxia-telangiectasia. Neurology. 1984;34:187-191. 6. Plaitakis A, Nicklas WJ, Desnick RJ. Glutamate dehydrogenase deficiency in three patients with spinocerebellar syndrome. Ann Neurol. 1980;
9. Obata K. Association of GABA with cerebellar Purkinje cells: a single cell analysis. In: Roberts E, Chase T, Tower D, eds. GABA in Nervous System Function. New York, NY: Raven Press; 1980:217-222. 10. Storm-Mathisen J. Distribution of the components of the GABA system in neuronal tissue: cerebellum and hippocampus\p=m-\effectsof axotomy. In: Roberts E, Chase T, Tower D, eds. GABA in Nervous System Function. New York, NY: Raven Press; 1976:149-168. 11. Nadi NS, McBride WJ, Aprison MH. Distribution of several amino acids in regions of the cerebellum of the rat. J Neurochem. 1977;28:453\x=req-\
7:297-303.
455.
12. Roffler-Tarlov S, Sidman RL. Concentrations of glutamic acid in cerebellar cortex and deep nuclei of normal mice and weaver, stagger, and nervous mutants. Brain Res. 1978;142:269\x=req-\ 283. 13. Nadi NS, Kanter D, McBride WJ. Effects of 3-acetylpyridine on several putative neurotransmitter amino acids in the cerebellum and medulla of the rat. J Neurochem. 1977;28:661-662. 14. Kish SJ, Perry TL, Hornykiewicz O. Increased GABA receptor binding in dominantly inherited cerebellar ataxias. Brain Res. 1983; 269:370-373. 15. Blass JP, Gibson GE: Studies in pathophysiology of pyruvate dehydrogenase deficiency. In: Kark RAP, Rosenberg RN, Schut LJ, eds. The Inherited Ataxias. New York, NY: Raven Press; 1978:181-194. 16. Kish SJ, Currier RD, Schut L, Perry TL, Morito CL. Brain choline acetyltransferase reduction in dominantly inherited olivopontocerebellar atrophy. Ann Neurol. 1987;22:272-275. 17. Hallak M, Giacobini E. Relation of brain
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
regional physostigmine concentration to cholinesterase activity and acetylcholine and choline
levels in rat. Neurochem Res. 1986;2:1037-1048. 18. Kark RA, Blass JP, Spence MA. Physostigmine in familial ataxias. Neurology. 1977;27:70\x=req-\ 72. 19. Manyam NVB, Hare TA, Katz L. Effect of isoniazid on cerebrospinal fluid and plasma GABA levels in Huntington's disease. Life Sci.
1980;26:1303-1308.
20. Deutsch SI, Mohs RC, Levy MI, et al. Acetylcholinesterase activity in CSF in schizophrenia, depression, Alzheimer's disease, and normals. Biol Psychiatry. 1983;18:1363-1373. 21. Grossman MH, Hare TA, Manyam NVB, Glaeser BS, Wood JH. Stability of GABA levels in
CSF under various conditions of storage. Brain Res. 1980;182:99-106. 22. Lundquist C, Blomstrand C, Hamberger A, Wikkelso C. Liquid chromatographic separation of cerebrospinal fluid amino acids after precolumn fluorescence derivatization. Acta Neurol Scand. 1989;79:273-279. 23. Ferraro TN, Hare TA. Triple-column ionexchange physiological amino acid analysis with fluorescent detection: baseline characterization of human CSF. Anal Biochem. 1984;143:82-94. 24. Johnson CD, Russell RL. A rapid, simple radiometric assay for cholinesterase. Anal Biochem. 1975;64:229-238. 25. Elble R, Giacobini E, Scarsella GF. Cholinesterases in cerebrospinal fluid: a longitudinal study in Alzheimer disease. Arch Neurol.
1987;44:403-407.
26. McCaman RE, Stetzler J. Radiochemical acetylcholine: modification for subpicomoles measurements. J Neurochem. 1977; 28:669-671. 27. SAS User's Guide: Statistics. Cary, NC: SAS Institute; 1982. 28. Perry TL, Berry K, Hansen S, Diamond S, Mok C. Regional distribution of amino acids in human brain obtained at autopsy. J Neurochem. assay for
1971;18:513-519. 29. Perry TL, Sanders HD, Hansen S, Lesk D, Kloster M, Graslin L. Free amino acids and related compounds in five regions of biopsied cat brain. J Neurochem. 1972;19:2651-2656. 30. Shaw RK, Heine JD. Ninhydrin positive
substances present in different areas of normal rat brain. J Neurochem. 1965;12:151-155. 31. Tallan HH, Rassin DK, Starman JA, et al.
Methionine metabolism in the brain. In: Lajtha A1 ed. Handbook of Neurochemistry. 2nd ed. New York, NY: Plenum Press; 1983;3:535-538. 32. Cohen SM, Nichols A, Wyatt R, Pullim W. The administration of methionine in chronic schizophrenic patients: a review of ten studies. Br J Psych. 1974;8:209-225. 33. Hamberger A, Berthold A, Jacobson I, et al. In vivo brain dialysis of extracellular nontransmitters on putative transmitter amino acids. In: Bayon A, Drucker-Colin R, eds. In Vivo Perfusion and Release of Neuroactive Substances. Orlando, Fla: Academic Press Inc; 1985:119-139. 34. Lehman A, Hamberger A. A possible interrelationship between extracellular taurine and phosphoethanolamine in the hippocampus. J Neurochem. 1984;42:1286-1290. 35. Oldendorf WH. Saturation of blood brain barrier transport of amino acids in phenylketonuria. Arch Neurol. 1973;28:45-48. 36. Manyam BV, Katz L, Hare TA, Kaniefski K, Tremblay RD. Isoniazid induced elevation of CSF GABA levels and effects on chorea in Huntington's disease. Ann Neurol. 1981;10:35-37. 37. Burgen ASV, Chipman LM. Cholinesterase and succinic dehydrogenase in the central nervous system of the dog. J Physiol. 1951;114:296-305. 38. Austin L, Phillis JW. The distribution of cerebellar cholinesterase in several species. J Neurochem. 1965;12:709-717. 39. Kasa P, Silver A. The correlation between choline acetyltransferase and acetylcholinesterase activity in different areas of the cerebellum of rat and guinea pig. J Neurochem. 1969;16:389\x=req-\ 396. 40. Feldberg W, Vogt M. Acetylcholine synthesis in different regions of the central nervous system. J Physiol. 1948;107:372-381. 41. Hebb CO, Silver A. Choline acetylase in the central nervous system of man and some other mammals. J Physiol. 1956;134:718-728. 42. Goldberg AM, McCaman RE. A quantitative microchemical study of choline acetyltransferase and acetylcholinesterase in the cerebellum of several species. Life Sci. 1967;6:1493-1500. 43. Kasa P. The cholinergic systems in brain and spinal cord. Prog Neurobiol. 1986;26:211-272. 44. MacIntosh FC. The distribution of acetylcholine in the peripheral and the central nervous system. J Physiol (Lond). 1941;99:436-442. 45. Welsh JH, Hyde JE. The distribution of acetylcholine in brains of rats of different ages.
J Neurophysiol. 1944;7:41-49. 46. Vizi SE, Palkovits M. Acetylcholine content in different regions of the rat brain. Brain Res Bull. 1978;3:93-96. 47. Kasa P, Banasghy K, Rakonczay Z, Gulya K. Postnatal development of the acetylcholine system in different parts of the rat cerebellum. J Neurochem. 1982;39:1726-1732. 48. Kasa P. The cholinergic systems in brain and spinal cord. Prog Neurobiol. 1986;26:211-272. 49. Thal LJ, Masur DM, Fuld PA, Sharpless NS, Davies P. Memory improvement with oral physostigmine and lecithin in Alzheimer's disease. In: Katzman R, ed. Banbury Report 15: Biological Aspects of Alzheimer's Disease. Cold Spring Harbor, Mass: Cold Spring Laboratory; 1983:461-469. 50. Zeisel SH, Growdon JH, Wurtman RJ, Magil SG, Logue M. Normal plasma choline responses to
ingested lecithin. Neurology. 1980;
30:1226-1229. 51. Blusztajn JK, Wurtman RJ. Choline and cholinergic neurons. Science. 1983;221:614-620. 52. Manyam BV. Recent advances in the treatment of cerebellar ataxias. Clin Neuropharmacol.
1986;9:508-516.
53. Barbeau A. Emerging treatments: replacetherapy with choline or lecithin in neurological diseases. Can J Neurol Sci. 1978;5:157-160. 54. Lawrence CM, Millac P, Stout GS, Ward JW. The use of choline chloride in ataxic disorders. J Neurol Neurosurg Psychiatry. 1980;43:452\x=req-\ 454. 55. Legg NJ. Oral choline in cerebellar ataxia. BMJ. 1978;2:1403-1404. 56. Legg N. Oral choline in cerebellar ataxia. BMJ. 1979;2:133. 57. Livingstone IR, Mastaglia FL, Pennington RJT, Skilbeck C. Choline chloride in the treatment of cerebellar and spinocerebellar ataxia. J Neurol Sci. 1981;50:161-174. 58. Schested P, Lund HI, Kristensen 0. Oral choline in cerebellar ataxia. Acta Neurol Scand. ment
1980;62:124-126. 59. Kark RA, Budelli MMR, Wachsner R. Double-blind triple-cross trial of low doses of oral physostigmine in inherited ataxia. Neurology. 1981;31:288-292. 60. Budelli MMR, Kark RAP, Blass JP, Spence
MA. Action of physostigmine of inherited ataxias. Adv Neurol. 1978;21:195-202.
Downloaded From: http://archneur.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/10/2015
