Movement Disorders Vol. 13, No. 3, 1998, pp. 522-528 0 1998 Movement Disorder Society

Monoaminergic Effects of High-Dose Corticotropin in Corticotropin-Responsive Pediatric Opsoclonus-Myoclonus

"Michael R. Pranzatelli, MD, TYung-yu Huang, MS, *Elizabeth Tate, MN, C, FNP, $David S. Goldstein, MD, PhD, $Courtney S. Holmes, MS, $Edward M. Goldstein, MD,

ItKaren Ketner, MSN, PNP, Morris Kinast, MD, ¶Bernadette M. Lange, MD, **Alfred Sanz, MD, ??Michael I. Shevell, MD, CM, FRCPC, Richard E. Stanford, MD, and $$Ingrid P. Taff, MD

"National Pediatric Myoclonus Center, Washington, DC; ?Department of Neuroscience, New York State Psychiatric Institute, and Department of Psychiatiy, Columbia University, New York, NY; $Clinical Neuroscience Branch, NINDS, NIH, Bethesda,

MD; $Scottish Rite Children's Center, Atlanta, GA; "Department of Pediatrics, UCSF Stanford Medical Center, Palo Alto, CA; ¶Division of Child Neurology, Arkansas Children's Hospital, Little Rock, AR; "*Department of Pediatrics, St. Joseph's Hospital,

Paterson, NJ; f fMontreal Children's Hospital, Montreal, Quebec; and $$Department of Pediatrics, Baptist Medical Center, Oklahoma City, OK, U.S.A.

Summary: Children with the opsoclonus-myoclonus syn- drome (OMS) usually respond to corticotropin (adrenocortico- trophic hormone, ACTH) treatment but the mechanism of ben- efit is unknown. We previously showed that both cerebrospinal fluid (CSF) homovanillic acid (HVA) and 5-hydroxyindole- acetic acid (5-HIAA) concentrations are low in pediatric OMS. In this study, we measured levels of CSF Dopa, catechol- amines, deaminated metabolites of catecholamines, as well as HVA and 5-HIAA in eight patients before and during treatment with ACTH. All the children were ACTH-responsive with 50- 70% improvement in multiple clinical features of OMS. ACTH treatment reduced the HVA concentration in every child by a mean of 21% (p < 0.001). Treatment with ACTH was associ- ated with significant correlations between dopaminergic mark-

ers such as HVA, dihydroxyphenylacetic acid (DOPAC), and Dopa. There were no significant changes in the CSF concen- trations of the noradrenergic markers norepinephrine (NE) and dihydroxyphenylglycol (DHPG), or the serotonergic marker 5- HIAA. The only child with a marked inflammatory pattern in CSF, which was reversed by ACTH, was atypical for a large increase in NE and decrease in 5-HIAA during ACTH treat- ment. Beneficial effects of ACTH in OMS are not associated with normalization of HVA or 5-HIAA levels. The pattern of decreased HVA and unchanged DOPAC levels could reflect decreased extraneuronal uptake of catecholamines (which ste- roids inhibit) or decreased 0-methylation of catecholamines in nonneuronal cells. Key Words: ACTH-CSF 5-HIAA- HV A-M yoclonus-Opsoclonus-OMS-Paraneoplastic.

Corticotropin (adrenocorticotrophic hormone, ACTH) is the principal pharmacologic treatment for childhood opsoclonus-myoclonus syndrome (OMS), a rare neuro- immunologic syndrome that also includes ataxia, dysar- thria, behavioral and cognitive problems. ' The mecha- nism of ACTH effectiveness is unknown, but ACTH has many adrenal and extra-adrenal effects, one of which may be to alter central neurotransmission.* We found low levels of the deaminated serotonin metabolite, 5- hydroxyindoleacetic acid (5-HIAA), and the deaminated,

Received June 12, 1997; revision received November 17, 1997. Accepted January 12, 1998.

Address correspondence and reprint requests to Michael R. Pranzatelli at the National Pediatric Myoclonus Center, 1828 L St., NW, Ste. 402, Washington, DC 20036, U.S.A.

0-methylated dopamine metabolite, homovanillic acid (HVA), in children with OMS compared with age- and gender-matched control^.^ These results suggest a distur- bance of brain monoaminergic function in some children with OMS, as in other myoclonic disorders4; however, the relationship of these alterations to levels of other compounds related to monoaminergic function and ef- fects of ACTH treatment are unknown.

We have hypothesized that a select population of neu- rons, perhaps sharing a common neurotransmitter phe- notype, may be the target of autoimmune-mediated brain injury in OMSs Neuroblastomas and ganglioneuroblas- tomas, found in approximately half of the children, con- tain monoamine receptors and metabolic enzymes6,' which could provide the antigenic cross-reactivity re- quired by this hypothesis and relevant to monoamine

522

ACTH IN OPSOCLONUS-MYOCLONUS 523

studies. In viral etiologies of OMS, which account for most of the other cases, cross-reactivity of viral coat proteins with neural antigens through molecular mimicry has been p r o p ~ s e d . ~ As an alternative to direct immuno- logic effects on central neurotransmitter systems, mono- aminergic neurotransmission could be disrupted second- arily through projections from a different, primary site of neural injury. In either scenario, measurement of a panel of other monoamine markers before and during ACTH treatment may be useful for understanding the neuro- chemical basis of the CSF abnormalities we found.

We report here the effects of ACTH treatment on the clinical findings and CSF levels of monoamines, Dopa precursors, and the catecholamine and monoamine me- tabolites in a comparatively large series of children with OMS. Besides 5-HIAA and HVA, we assayed dopamine (DA), the catecholamine precursor dihydroxyphenylala- nine (Dopa), dihydroxyphenylacetic acid (DOPAC) (the deaminated metabolite of DA), and dihydroxyphenylgly- col (DHPG) (the deaminated metabolite of norepineph- rine [NE]).

METHODS

Subjects All children were evaluated through the National Pe-

diatric Myoclonus Center, were moderate to severely af- fected and required treatment, and were enrolled consec- utively after informed consent was signed for this Insti- tutional Review Board-approved protocol. Gender was evenly matched, and the age range was 1.6-4.4 years (mean 2.5 rt 0.3 years, standard error of mean [SEMI). One case was paraneoplastic (neuroblastoma) and the others were paraviral. Although paraneoplastic and para- viral etiologies of OMS occur with nearly equal fre- quency, more patients with paraviral etiology were avail- able by chance for enrollment into the study. A thorough investigation for an occult neuroblastoma, including an [ '231]MIBG (meta-iodobenzyl guanidine) scan, urinary catecholamines, body CT scan, neuron-specific enolase, and serum ferritin, was done. All patients were seroneg- ative for the paraneoplastic autoantibodies anti-Hu, anti- Ri, and anti-Yo, as is common in pediatric OMS.5 In nonparaneoplastic cases, viral titers, serum biotin and biotinidase, antinuclear antibody, and routine tests, such as complete blood count, electrolytes, and liver function tests, were also done.

Eight patients were evaluated before treatment and after 2 months on ACTH (40 IU/cc, Acthar gel admin- istered intramuscularly; Rh8ne-Poulenc Rorer Phanna- ceuticals, Collegeville, PA, U.S.A.). The ACTH dosing schedule was a modification of a high-dose schedule for

infantile spasms,8 starting with 75 IU/m2 administered intramuscularly twice a day for I week, then daily for I week, every other day for 2 weeks, then tapered over 16 weeks (70, 60, 50, 45, 35, 25, 15, 10, 8 IU/m2) on alter- nating days, remaining at the lowest dose for 4 weeks. At the time of the second lumbar puncture, the patients were receiving approximately 60 IU/m2 (range, 45-75 IU/m2) on alternate days. This time point was selected to allow enough time for maximal clinical response while avoid- ing relapses on lower doses.

Each child was videotaped before each lumbar punc- ture, and opsoclonus, myoclonus, ataxia, irritability, and functional disability were scored from the videotape by a trained observer blinded to treatment status using the following scale: 0 = absent, 1 = mild, 2 = moderate, and 3 = severe. Functional disability was an overall rating based on impairment of activities such as walking, reaching for objects, and ability to interact with the ex- aminer. The scoring method was validated previously through the use of a judge panel. Videotaping technique, blinding procedures, quality control issues, scale valida- tion methods, and the advantages and disadvantages of rating scales and videotape-based scoring systems in children with myoclonus have been discussed previ~usly.~

CSF Studies Lumbar punctures were performed at the L4-5 inter-

space in the lateral decubitus position in the morning after overnight fasting and the fourth and fifth milliliter aliquots of cerebrospinal fluid removed were used for neurochemical analysis. The children required sedation, which was usually propofol or midazolam, and the same sedative was given for both lumbar punctures. We have found no significant effect of these sedatives on CSF HVA or S-HIAA.3

Assays for CSF 5-HIAA and HVA were performed by HPLC with ESA Coulochem electrochemical detection using a modification3 of the method of Kalivas and Bron- son." For measurement of CSF catecholamines, Dopa, or metabolites, a batch alumina extraction method was used. Intraassay and interassay variability have been evaluated previously and the absolute value of CSF monoaminergic markers is comparable among series and

The samples were run in batches with CSF standards in all assays.

Concentration differences between the various mono- aminergic markers in CSF affect detectability. In our pediatric CSF samples, HVA and 5-HIAA concentra- tions are 30- to 50-fold higher than are DHPG, DOPAC, and Dopa; 200- to 300-fold more than NE; and 2000- to 5000-fold more than DA. The rank order is HVA > S- HIAA >> DHPG > DOPAC, Dopa >> NE > DA. Be-

Movement Disorders, Vol. 13, No. 3, 1998

524 M. R. PRANZATELLI ET AL.

J

* 2- c, -I s K 0 z m a 1-

0-

cause the concentration of dopamine is low and near the limit of detectability, it would not be possible using this assay to demonstrate a treatment-induced reduction in dopamine.

Although patients with OMS served as their own con- trols, we also calculated CSF normative data for HVA and 5-HIAA from our larger da taba~e .~ We have previ- ously shown that CSF HVA and 5-HIAA levels are cor- related inversely with age and adult levels are achieved by approximately 4 years of age.3 CSF HVA and 5- HIAA concentrations from age- and gender-matched un- treated children with OMS from the database were also compared to assess the validity of pretreatment values because the sample size in the current study is small.

Statistics Statistical analysis was performed using general linear

models (PROC GLM) of the Statistical Analysis System (SAS), a statistical computer software package.13 Statis- tical tests included paired or unpaired t tests, Pearson correlations, and analysis of variance (ANOVA).

RESULTS

Clinical Response ACTH treatment significantly reduced neurologic ab-

normality in children with OMS (p = 0.001, ANOVA). There was no significant difference in response of indi- vidual syndrome component features to ACTH. The

T 31

OPS MY0

T

ATX

mean improvement in severity scores was 70% for ops- oclonus, 52% for myoclonus, 50% for ataxia, 55% for irritability, and 57% for overall functional disability. All the children responded significantly to ACTH regardless of etiology of OMS, decreasing in severity from the moderate to severe abnormality range to the mild to moderate range (Fig. 1). Improvement was global and sometimes dramatic, but no patient became completely normal.

CSF Protein and Cells The mean protein concentration was higher in CSF

before ACTH treatment (51 f 28, SEM) than during ACTH treatment (20 k 3 mg/dL). This difference was the result of two patients, one whose protein was 213 before treatment (RBC 350, WBC 25) and 30 (RBC 945, WBC 2) during treatment (subsequently referred to as the out- lier), and another whose protein was 53 (RBC 910, WBC 18) and 15 (RBC 900, WBC 4). All other patients had normal CSF protein levels. Mean CSF white blood cell counts/cm2 were 8.4 k 3 (range, 1-25) and 2.5 k 0.8 before and during treatment with ACTH, respectively. The corresponding mean CSF glucose values were 50 k

5 and 54 k 6 mg/dL.

Dopaminergic Markers The most consistent change in CSF neurochemistry

associated with ACTH treatment was a 21% reduction in the concentration of CSF HVA. Reductions were found

T

I R R D IS

f 0 PRE

FIG. 1. Effect of ACTH treatment on the clinical features of OMS about the time the second lumbar puncture was performed compared with pretreatment evaluation (PRE). Patient videotapes were scored by an observer blinded to treatment status using a scale of abnormality from &3 for each clinical feature of OMS: opsoclonus (OPS), myoclonus (MYOj, ataxia (ATXj, irritability (IRRj, and overall functional disability (DIS). Data are means f SEM. Asterisks indicate statistical significance on paired t tests, p = 0.001-0.004.

Movemenf Disorders, Voi. 13. No. 3, 1998

ACTH IN OPSOCLONUS-MYOCLONUS 525

CI

E . c Y

CI - E a . c - P

A

'"1 100 7

"a P 0

PRE ACTH

a reduction of 2049% which was offset by increases in two patients, including an increase of 57% in one of them (Fig. 2B). The largest reductions in CSF 5-HIAA con- centrations during ACTH treatment occurred in the two patients with the highest pretreatment 5-HIAA values.

Correlations of Monoaminergic Markers The effect of ACTH treatment on the relation of dif-

ferent monoaminergic markers was evaluated by corre- lation analysis. Before treatment, there were few positive statistical correlations between CSF levels of the mea- sured neurochemicals (Table l). CSF 5-HIAA and HVA were positively correlated, but DOPAC and HVA, Dopa and HVA, and CSF NE and DA were not. During ACTH treatment, several changes were found: the dopaminergic markers DOPAC and HVA, Dopa and DOPAC, and Dopa and HVA became significantly correlated. As ex- pected, the positive correlation between 5-HIAA and

' O 1 0 ACTH

PRE

FIG. 2. Effect of ACTH treatment on the concentration of CSF HVA (A) and 5-HIAA (B) compared with pretreatment values (PRE). Pa- tients were their own controls. The range in pretreatment levels is the result mainly of the inverse correlation of patient age and CSF HVA and 5-HIAA concentrations.

in every patient (Fig. 2A) and ranged from 843%. DOPAC levels did not change significantly. Although there was a trend toward lower levels of the DA precur- sor Dopa in the ACTH-treated group, there was no con- sistent effect on the concentration of dopamine (Fig. 3).

Noradrenergic Markers There were no significant changes in CSF NE Or FIG. 3. Group comparison of concentrations of monoamipe neuro-

transmitters, precursors or metabolites before (PREJ and during ACTH treatment. Data are means f SEM. Asterisks Indicate statistical signifi- cance on paired t tests (p = 0.007). Our CSF normative values in control subjects younger than 4 years old (n = 30) were 99.2 i 5.9 ng/mL (range, 60-183) for HVA and 51 f 6 n@mL (range, 17-98) for 5-HTAA.

DHPG levels with ACTH treatment.

Serotonergic Markers There were no significant changes in mean CSF 5-

HIAA during treatment with ACTH. Six patients showed

Movement Disorders, Vol. 13, No. 3, 1998

526 M. R. PRANZATELLI ET AL.

TABLE 1. Effect of ACTH treatment on correlations of CSF monoarninergic murkers

~

Markers compared r P

5-HIAA and HVA Pretreatment On ACTH

DOPAC and HVA Pretreatment On ACTH

Dopa and HVA Pretreatment On ACTH

Pretreatment On ACTH

DHPG and Dopa Pretreatment On ACTH

DOPAC and Dopa Pretreatment On ACTH

NE and DA

0.82 0.85

0.43 0.89

0.02 0.75

0.41 0.99t

0.49 0.66

0.49 0.85

0.01* 0.007*

0.29 (ns) 0.003*

0.97 (ns) 0.03"

0.15 (ns) 0.0001 t

0.21 (ns) 0.08 (ns)

0.21 (ns) 0.007*

Asterisks indicate statistically significant (p < 0.05) correlations (r) between CSF monoaminergic markers (PROC CORR). ns = not sig- nificant.

t With outlier excluded, r = 0.69, p = 0.09 on ACTH.

HVA, which probably has to do with the pharmacologic relation of 5-HT and DA neuronal system^,^ was not altered by ACTH treatment.

Ratios of Monoaminergic Markers Various ratios were calculated as indices of turnover

or synthesis of monoaniine neurotransmitters (Table 2). Unfortunately, the ratio of HVA:DA, which would have been most relevant for turnover, and Dopa:DA could not be computed in each case because DA was often unde- tectable. ACTH did not induce significant changes in ratios. This may have been the result of the wide range of values.

An Outlier The only child with markedly elevated CSF protein

with pleocytosis on the first lumbar puncture had a unique pattern of changes in CSF neurochemistry. Whereas the effect of ACTH on dopaminergic markers was similar to other patients, there were also effects on noradrenergic and serotonergic markers. This was not a technical artifact of elevated CSF protein because CSF samples were deproteinated by addition of trichloroace- tic acid and centrifugation before assays, the pellet was discarded, and only the supernatant was used for the assay.

Unlike other patients during ACTH treatment, norepi- nephrine increased fivefold and dopamine increased 16- fold, and there was a 50% increase in DOPAC and a 70%

increase in DHPG. ACTH treatment was associated with decreases in the ratios of HVA:DA (22-fold), Dopa:DA ratio (20-fold), D0PAC:DA (1 1-fold), HVA:DOPAC (twofold), a twofold increase in DHPG:NE ratio, and no change in HVA:Dopa ratio. Pretreatment CSF HVA and 5-HIAA values were in the high normal range according to the 1iteratu1-e'~ and our control value^.^

Similar to some other patients during ACTH treat- ment, CSF Dopa, HVA, and 5-HIAA were reduced by 17%, 22%, and 40%, respectively.

When statistical analysis was performed with and without the outlier in the data, the outlier altered statis- tical significance only in two instances. The lowering effect of ACTH on CSF 5-HIAA just reached signifi- cance (p = 0.048, paired t test) when the outlier was excluded. In the second instance, only with the outlier included were NE and DA significantly correlated during ACTH treatment.

DISCUSSION At a time of 50-70% clinical improvement during

ACTH treatment in children with OMS, the CSF HVA concentration was 21% decreased. CSF HVA concentra- tions are already 3040% lower in young children with OMS than in age- and gender-matched control subject^.^ These data suggest that ACTH treatment in pediatric OMS is associated with a reduction in brain HVA, be- cause CSF and brain HVA levels are positively corre- lated in the human brain.'* The ACTH-associated corre- lations of dopamine precursors and metabolites further support changes in the dopaminergic system. Although CSF 5-HIAA is also low in OMS,3 ACTH did not sig- nificantly affect its concentration. Whether ACTH acts directly on the brain as ACTH,.,, or an active cleavage fragment or has an indirect effect mediated by cortisol or some other intermediate is unknown.2

There is precedent for ACTH lowering HVA but not 5-HIAA. In infantile spasms, which has been reviewed re~ent ly , '~ ACTH treatment significantly reduced CSF

TABLE 2. Effect of ACTH treatment on ratios of CSF monoaminergic markers

~~ ~ ~ _ _ _ _ _ ~

Marker ratio Pretreatment On ACTH

HVA: Dopa 58.3 f 14 (31-79) 52.3 i 19 (25-80) HVA:DOPAC 55.6 2 8.4 (45-66) 47.0 ? 21 (20-80) DHPG:NE 10.4 f 6.3 (5-24) 9.7 k 7.5 (2-27) D0PAC:Dopa 1.02 k 0.3 (0.69-1.58) 1.17 ? 0.4 (0.72-1.86) DHPG:Dopa 1.77 f 0.6 (0.91-2.53) 1.68 i 0.5 (1.07-2.63)

Data are means + standard error of mean. The range of values is stated in parentheses. ACTH had no significant effect on the ratios of CSF markers.

Movement Disorderv, Vul. 13, Nu. 3, 1998

ACTH IN OPSOCLONUS-M YOCLONUS 52 7

HVA by 40% in two of three studies, whereas it had no significant effects on 5-HIAA in three studies and in- creased in NE in one study (possibly relevant to our outlier). Pretreatment CSF HVA in infantile spasms has been reported as high, low, or normal. In one of these studies, ACTH further lowered an already low level of HVA in CSF.

One interpretation of the apparent further lowering effect of ACTH on CSF HVA in OMS is that low CSF concentrations in OMS reflect compensatory mecha- nisms rather than primary neurologic dysfunction and ACTH enhances the compensatory mechanisms. It is also possible that ACTH exerts or accelerates a matura- tional effect on the dopaminergic system as CSF HVA levels decrease with age in normal control subjects. An- other interpretation is that the reduction by ACTH is the result of regression to the mean: not real changes in what is being measured but rather a tendency of high values to regress to the mean. Three of the children in the present study had higher CSF HVA and 5-HIAA levels than we have found previously. The mean pretreatment CSF HVA level in this study (86.4 k 7.9 ng/mL) was higher but was not significantly different than levels in age- matched children with OMS (n = 14) who were taking ACTH or steroids at the time of lumbar puncture taken

from our database (65.1 5 19.5 ng/mL).3 The same is true of CSF 5-HIAA: 38.2 +- 6.9 ng/mL in this study com- pared with 27.9 _+ 8.0 ng/mL in our d a t a b a ~ e . ~ Against regression to the mean as an explanation, however, is the reduction of every CSF HVA value, not just the mean, during ACTH treatment. Aside from the issue of which, if any, of these interpretations is correct, the main point is that beneficial effects of ACTH in OMS were not associated with the expected effect, which was normal- ization of HVA or 5-HIAA levels.

The effects of ACTH on central monoaminergic mark- ers in pediatric OMS are compatible with effects of ACTH treatment on monoaminergic systems in healthy experimental animals, which vary with brain region, ACTH dose, and duration of ACTH treatrnenL2 ACTH increases DA release in septa1 slices, but not in striatum or nucleus accumbens, and increases DA synthesis in cortex. ACTH increases NE synthesis or turnover in the brain stem, hypothalamus, and cortex in some but not other studies. The sequential metabolism of tyrosine to Dopa, dopamine, and then norepinephrine involves tyro- sine hydroxylase, Dopa-decarboxylase, and dopamine-b- hydroxylase (Fig. 4). ACTH treatment increases activity of tyrosine hydroxylase in the locus ceruleus. l6 Steroids also inhibit nonneuronal uptake of catecholamines. Ef-

TYRH LAAD DBH A

Tyrosine - DOPA - DA ----+ NE MA0 1 1 ARD

DHPG DOPAC I COMT 1 COMT

MHPG HVA

B TRPH LAAD

Tyrosine -- 5-HTP 5-HT

5-HIAA FIG. 4. Schema of monoamine metabolism and possible sites of modulation by ACTH. For simplicity, the pathways shown indicate the main but not the only routes of metabolism. (A) Enzymes for catecholamine metabolism indicated are tyrosine hydroxylase (TYRH), L-aromatic amino acid decarboxylase (same as dopa decarboxylase), dopainine-beta-hydroxylase (DBH), monoamine oxidase (MAO), catechol-0-methyltransferase (COMT). aldehyde dehydxogenase (ADH), and aldehyde reductase (ARD). The alternative pathway of generating norepinephrine from dopamine is active only in noradrenergic cells containing DBH. (B) Enzymes for serotonin metabolism are tryptophan hydroxylase (TRPH), LAAD, MAO, and ADH. A drop in CSF HVA but not DOPAC during ACTH treatment suggests decreased activity of COMT. To explain the lower levels of both HVA and 5-HIAA in OMS by a shared metabolic pathway, decreased activity of M A 0 could be postulated.

Movement Dis0rder.v. V d 1 7 Nn 9 IOOR

528 M. R. PRANZATELLI ET AL.

fects of ACTH on 5-HT include decreased concentra- tions and turnover in cortex and hindbrain, increased levels in hippocampus and hypothalamus, or no changes in other studies. The fall of CSF HVA without a drop in DOPAC during ACTH treatment suggests decreased ac- ti vity of catechol-0-rnethyltran sferase (COMT). De- creased activity of monoamine oxidase (MAO) should reduce levels of both HVA and 5-HIAA. Downregula- tion of DA receptors or 5-HT receptors by ACTH2 would also explain the reduction in HVA and 5-HIAA.

The role of monoamines in the clinical effect of ACTH in OMS remains speculative. ACTH also could improve OMS by other mechanisms. Because all chil- dren responded to ACTH treatment, it is not possible to conclude that neurochemical changes in ACTH-treated children were responsible for clinical improvement. Mono- aminergic effects of ACTH may be secondary to ACTH- induced improvement in OMS symptoms rather than ACTH itself. Whether resolution of OMS without treat- ment is also accompanied by decreased HVA is unknown because severe OMS does not resolve spontaneously.

We reported the outlier in case it represents a bio- chemical subgroup of OMS. This possibility can not be excluded in a small sampling of a rare disorder. Because it is likely that new immunologic markers will be de- scribed in pediatric OMS, the description of biochemical subgroups may aid in clinical-biochemical-serologic correlation. Another possibility is that the higher pre- treatment CSF 5-HIAA in this patient may have been related to the apparent inflammatory process. The CSF pleocytosis and greatly elevated protein concentration, a less common finding in OMS,' might indicate greater immunologically mediated injury. ACTH treatment, which normalized the CSF protein and cell count, has been reported to reduce protein leakage into the CSF in infantile spasm^.'^

Little information is available on CSF levels of Dopa, catecholamines, and deaminated metabolites of catechol- amines in children. The values for CSF DOPAC and Dopa we found in untreated children with OMS would be high for adults but DA, NE, and DHPG concentra- tions would not.'' A study to evaluate CSF DA, NE, DOPAC, Dopa, and DHPG in pediatric control subjects is in progress at our Center.

Acknowledgments: This work was supported in part by grants to Dr. Pranzatelli and E. Tate from the Food and Drug Administration Orphan Products Research and Development (FD-U-000746 and FD-U-000955). The authors thank Gail Cain for typing the manuscript.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Pranzatelli MR. The neurobiology of opsoclonus-myoclonus. Clin Neuropharmacol 1992;15: 186-228. Pranzatelli MR. On the molecular mechanism of adrenocorticotro- phic hormone: neurotransmitters and receptors. Enp Neurol 1994;

Pranzatelli MR, Huang Y-Y, Stanley M, et al. Cerebrospinal fluid 5-hydroxyindoleacetic acid and homovanillic acid in pediatric opsoclonus-myoclonus. Ann Neurol 1995;37: 189-197. Pranzatelli MR. Serotonin and human myoclonus: rationale for the use of serotonin receptor agonists and antagonists. Arch Neurol 19943 1 :605-617. Pranzatelli MR. The immunophmacology of the opsoclonus- myoclonus syndrome. Clin Neuropharmacol 199639: 1 4 7 . Pranzatelli MR, Balletti J. Identification of 5-HTlA recognition sites in buman ganglioneuroblastoma. Eur J Pharmacul 1989; 170:

Pranzatelli MR, Balletti J. Serotonin receptors in human neuro- blastoma: a possible biologic tumor marker. Exp Neurol 1992;l IS: 423427. Snead OC, Benton JW, Hosey LC, et al. Treatment of infantile spasms with high-dose ACTH: efficacy and plasma levels of ACTH and cortisol. Neurology 1989;39: 1027-1031. Tate E, Pranzatelli MR, Ho H, Kim A. A sensitive and semiquan- titative pediatric myoclonus evaluation scale. J Neurosci Nurs 1995;27:287-291. Kalivas PW, Bronson M. Mesolimbic dopamine lesions produce an augmented behavioral response to enkephalin. Neurophanna-

Eisenhofer G, Goldstein DS, Stull R, et al. Simultaneous liquid chromatographic determination of 3,4-dihydroxyphenylalanine in plasma and their responses to inhibition of monoamine oxidasc. Clin Chem 1986;32:2030-2033. Stanley M, Traskman-Bendz L, Dorovini-Ais K. Correlations be- tween aminergic metabolites simultaneously obtained from human CSF and brain. Life Sci 1985;37:1279-1286. SAS Institute Inc. SAS/STAT User's Guide, Release 6.03 edition. Cary, NC: SAS Institute Inc, 1988. Langlais PJ, Walsh FX, Bird ED, Levy HL. Cerebrospinal fluid neurotransmitter metabolites in neurologically normal infants and children. Pediatrics 1985;75:580-586. Pranzatelli MR. Putative neurotransmitter abnormalities in infan- tile spasms: CSF neurochemistry and response to drugs. J Child Neurol 1994;9: 1 19-129. Markey KA, Sze PY. Influence of ACTH on tyrosine hydroxylase activity in the locus coeruleus of mouse brain. Neuroendocrinology

Siemes H, Siegert M, Aksu F, Emrich R, Hanefeld F, Scheffner D. CSF protein profile in infantile spasms. Influence of etiology and ACTH on treatment. Epilepsia 1984;25:368-376.

125:142-161.

127-128.

Cology 1985;24:931-936.

1984;38:269-275.

Movement Disorders, Vol. 13, No. 3, 1998