 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Division of Neurology
General Clinical Research Center
|| Physical and Occupational Therapy
¶ University of Southern California University Affiliated Program at Childrens Hospital Los Angeles
Department of Neurology, Keck School of Medicine, University of Southern California, Los Angeles, California
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods.Thirteen of 17 children with opsoclonus-ataxia, all with neuroblastoma, who were previously reported were reevaluated a second time 2 to 4 years after the initial assessment. One subject who lived out of state was partially reevaluated and is included. Five new subjects (2 with neuroblastoma and 3 without) were also enrolled. Each was evaluated twice at a minimum interval of 1 year between sessions. Intercurrent medical course was recorded, emphasizing medication and relapse history. Cognitive, adaptive behavior, academic, speech and language, and motor abilities were assessed.
Results.For the group as a whole, overall standardized, age-adjusted cognitive scores improved. Generally, younger subjects' cognitive and adaptive behavior scores improved more than older subjects. Although all subjects had gains in speech, language, and motor function, some progressed at a slow pace, and in some instances, standard scores dropped. There was a striking influence of clinical course. Although initial presentation was severe and all subjects required high doses of corticosteroids or corticotropin, 5 had a monophasic course and were able to be weaned from treatment without relapses. Fourteen had multiple relapses over the years, generally with reduction of medication or intercurrent illnesses. Of the 5 children with monophasic course, 4 are currently functioning in the average range with a full-scale IQ of 
90 and age-appropriate academic and adaptive skills.
Conclusions.The results continue to raise concern that opsoclonus-ataxia is sometimes a progressive encephalopathy. A minority of children with opsoclonus-ataxia have a monophasic course. Despite initial severity of symptoms, these children may have a more benign prognosis. For the majority of children with opsoclonus-ataxia, the course includes multiple relapses and requires prolonged treatment. Developmental sequelae are significant in these children with chronic course.
Key Words: opsoclonus • neuroblastoma • outcome • development • paraneoplastic syndromes • ataxia
Abbreviations: ACTH, corticotropin • IVIg, intravenous immunoglobulin • VABS, Vineland Adaptive Behavior Scales • DQ, developmental quotient
Opsoclonus-ataxia (sometimes alternatively called opsoclonus-myoclonus syndrome, Kinsbourne syndrome, or dancing-eye syndrome), when seen in infants and toddlers, is often the presenting feature of occult neuroblastoma.1–5 Clinically indistinguishable opsoclonus-ataxia syndrome presents in children in the same age range, apparently without neuroblastoma, and is thought to be postinfectious. Adult paraneoplastic opsoclonus-ataxia has been described primarily in association with occult small-cell carcinoma of the lung and breast and ovarian carcinoma and is often associated with specific paraneoplastic antibodies, anti-Yo and anti-Hu.6–8
Extensive investigations of childhood opsoclonus-ataxia, paraneoplastic or not, have not uncovered a specific antibody or tumor antigen, although various autoantibodies have been described in isolated cases.9–13 The immunologic mechanism of childhood opsoclonus-ataxia is not completely clear.
Despite tumor resection, nearly all children with opsoclonus-ataxia require vigorous and prolonged immunosuppressive treatment to control the neurologic manifestations.14 A variety of agents and combinations are used, but nearly all children with opsoclonus-ataxia require moderate to high doses of corticosteroids or corticotropin (ACTH) for months to years. Additional treatment with intravenous immunoglobulin (IVIg),15–18 azathioprine, or other immune suppressants is often added when response is suboptimal or tapering off corticosteroids or ACTH is otherwise impossible. Evidence regarding a benefit of early use of chemotherapy for children with neuroblastoma-associated opsoclonus-ataxia has been mixed, with 1 study finding a significant benefit of chemotherapy in reducing neurologic sequelae19 and several others failing to confirm benefit.20, 21 Developmental sequelae are frequent and may be severe regardless of both underlying etiology and treatment.19, 20, 22–26
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Three of the new subjects did not have a neuroblastoma found. However, their clinical presentations and treatment were otherwise indistinguishable from the children with neuroblastoma, so they are included in this report. In addition to previously reported clinical data and treatment information, we paid particular attention to whether the symptoms were recurring or increasing with infections or with tapering of medication during the interval between tests.
We were specifically concerned with the question of whether developmental scores worsened over time for children with opsoclonus-ataxia and with identifying treatment or disease factors that differentiated between children showing progressive improvement from those with declining function. We hypothesized that children with frequent relapses of opsoclonus-ataxia symptoms as medications were tapered or with intercurrent illnesses would be more at risk for progressive decline or lack of age-related progression of neurocognitive function that those whose improvement was sustained without repeated setbacks.
After approval from the institutional review board at Childrens Hospital Los Angeles, parents of the original subjects residing in California were contacted and invited to have their child participate in a comprehensive reevaluation. After consent was obtained, each child's clinical information was abstracted from available records and parent report. All 13 of the original subjects who lived in California were available for reevaluation. Four subjects from other states were not included because of lack of funding for transportation. One out-of-state family had testing done locally and had Vineland Adaptive Behavior Scales27 (VABS) administered by telephone. Results from the VABS are included for this subject. Five new subjects were seen, 2 with neuroblastoma and 3 with opsoclonus-ataxia without evidence of tumor. Each was examined twice, with testing repeated at an interval of at least 1 year.
A clinical neurologic examination was performed. Over several test sessions, each child underwent a full neurocognitive and behavioral evaluation by a neuropsychologist (V.L.B.), a speech and language evaluation by a speech-language pathologist (S.K.A.), and an evaluation of motor skills by an occupational therapist (K.E.P.). When possible, the same test battery was used as in the prior evaluation, but because of age differences, this was not generally possible. The tests used are listed in Table 1. For purposes of analysis, full-scale IQ (from either the Wechsler Preschool and Primary Scale of Intelligence, Revised,28 or Wechsler Intelligence Scale for Children, 3rd edition29) and Mental Development Index (Bayley Scales of Infant Development-230) were converted to z scores and combined as a single measure of overall ability.
|
Motor testing was conducted by an occupational therapist, who reassessed 10 of the returning subjects and 4 of the new subjects. The Peabody Developmental Motor Scale34 was used for children <4 years of age, and the Bruininks-Oseretsky Test of Motor Proficiency for children 4 years of age.35 In younger children, the same test could not be used, because subjects outgrew the age range of the Peabody Developmental Motor Scale. In other cases, standard scores could not be calculated, because subjects were functioning at a level that was too low. Thus, for most subjects, change scores could not be calculated because of either test transitions between the first and second session or very low scores on 1 or both of the testing sessions.
Assessment of behavioral adjustment was based on the Achenbach Child Behavior Checklist.36 Two of the younger subjects whose parents were Spanish-speaking were assessed with the Bayley Behavior Rating Scale.30
Speech and language assessment was highly variable because of range of age and abilities of the subjects. Speech and language testing was conducted by the original speech-language pathologist, who reassessed 10 of the returning subjects and 4 of the new subjects. As in the original study, the Peabody Picture Vocabulary Test, 3rd edition,37 and the Expressive Vocabulary Test38 were used to assess single-word receptive and expressive vocabulary knowledge, respectively, except for 1 subject who was administered the Expressive One -Word Picture Vocabulary Test-Revised.39 The Test of Language Development, Primary and Intermediate, 3rd Edition,40, 41 were used to assess receptive and expressive language in 7 of the returning 10 subjects to accommodate the subjects' current chronological ages (5 years 4 months to 13 years 3 months), substituting for the originally used tests (ie, the Preschool Language Scale, 3rd Edition,42 and the Clinical Evaluation of Language Fundamentals-Preschool43). The listening and speaking quotients, each of which reflected performance on 2 subtests of the Test of Language Development, were used as measures of receptive and expressive language performance, respectively. The Preschool Language Scale was readministered to 2 of the returning subjects, 1 who was still within the test age range and 1 who was out of range because of her overall developmental level. The originally administered subtests of the Clinical Evaluation of Language Fundamentals were readministered to 1 of the 10 returning subjects. Although most of the returning subjects were chronologically out of range for the originally used developmental measures of spontaneous expressive language (ie, mean length of utterance in morphemes44 and Brown's 14 morphemes45), these were scanned again for changes in those subjects who had been severely delayed at the time of the original study. In addition, sentence-structure complexity observed in spontaneous language samples (ie, syntactic coordination and subordination) was also noted. Of the 4 new subjects tested, 1 completed all formal test administrations, whereas the remaining 3 were able to participate partially.
Speech articulation and speech intelligibility were assessed in spontaneous speech samples and/or by using the Goldman Fristoe Test of Articulation46 with those subjects who could participate in the picture-naming and scene-description tasks of this test.
In addition to clinical information collected at the first evaluation, we recorded clinical course during the interval between evaluations. We were particularly interested in 2 specific aspects of the child's clinical course in the interval between neurocognitive evaluations: treatment during that interval, with ACTH, corticosteroids, or other immunosuppressant medications, and relapses between evaluations, either with attempts to taper medication or with intercurrent illnesses. The third clinical point of interest was whether there were relapses at any point (either with intercurrent illness or tapering of medication) or whether the course was monophasic, with sustained improvement once treatment began, without setbacks.
Data were analyzed by using a standard statistical package, SAS/STAT 6.12 of the SAS system for UNIX (SAS Institute, Inc, Cary, NC). Because of small sample size, only a limited number of hypotheses were subjected to formal statistical analysis. Only the overall cognitive measure, academic achievement, overall adaptive behavior measures, and the VABS were analyzed statistically. All test scores were converted to z scores for purposes of analysis. A z score of 0 corresponds to the mean for the test, and the standard deviation (SD) is 1. Thus, a z score of 0 corresponds to an IQ or development quotient (DQ) of 100. Because virtually no subject could be tested at both sessions with the same speech and language measures or motor scales, these measures were only used descriptively.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Of the 19 subjects, 16 had neuroblastomas. Three had no tumor found despite extensive diagnostic testing. All neuroblastomas were "low risk" without N-myc amplification, and none have recurred. All but 1 were treated with surgery alone. One child received chemotherapy before definitive resection because of large tumor size.
Cognitive and Adaptive Abilities
In general, overall cognitive scores improved between testing sessions for the group as a whole, with a mean z score increase of .51. Individually, range was from a z score decrease of 1.3 to an increase of 2.4. Increase in IQ/DQ over the interval between test sessions was negatively correlated with age at the time of testing, but this did not reach statistical significance (ie, a larger average gain in IQ/DQ was seen with younger children). See Fig 1 for a graphic depiction of the relationship of first and second test scores with age.
|
|
Gross motor z scores at the second test session ranged from 0.7 to below –3.0. For 6 subjects for whom the gross motor testing change score could be calculated, change ranged from 1.55 to –4.03. The available change scores do not adequately describe the group, however. Four additional subjects could not have change scores calculated because standard scores were below –3.0 at 1 or both testing sessions. Of these 4, 1 had substantially improved into the scorable range, whereas 1 had worsened from a z score of –2.5 to <3.0, without any loss in raw score.
Behavioral and Emotional Adjustment
Assessment of behavioral and emotional adjustment at the second evaluation was based on mothers' responses to the Achenbach Child Behavior Checklist. Scores above z = 2.0 were in the clinically significant range and z scores between 1.0 and 2.0 were in the borderline range of clinical significance. Results indicated the following emotional or behavioral problems: of the 16 subjects who received behavioral assessments at the second test session, 7 had a total score in the borderline or clinically significant range (3 subjects in the clinical range and 4 in the borderline clinical range). Of the 7 subjects who had elevated total scores, 5 subjects had externalizing symptoms (4 subjects in the borderline range and 1 in the clinically significant range), and 5 had internalizing symptoms (4 in the borderline range and 1 in the clinically significant range). On the syndrome scales, clinically significant elevations were found on the following scales: anxious/depressed (n = 3); withdrawn (n = 4); somatic problems (n = 7); aggression (n = 4); attention problems (n = 5); social problems (n = 6); and thought problems (n = 5). Both of the younger subjects (<4 years of age) had significant sleep problems (n = 2).
Behavioral assessment of the 2 subjects who were evaluated at the second test session with the Bayley Behavior Rating Scale indicated that 1 child was in the normal range and the other was in the nonoptimal range on total score. Both children were in the normal range on the orientation/engagement scale. However, 1 child was in the normal range and the other in the questionable range on the emotional-regulation scale. Both were in the nonoptimal range on the motor-quality scale.
Compared with data gathered from our initial study, mothers' responses yielded a very small increase in the number of emotional and behavioral problems as indicated by scores on the total, internalizing- and externalizing-problems, anxiety, and somatic-problems scales. However, overall it seems that behavior and emotional functioning were stable and did not worsen significantly as the subjects got older. In fact, mothers' responses yielded a decrease in the number and severity of attention problems compared with data from the initial study.
Language Assessment
Results of speech and language testing at the second test session continued to show deficits in a majority of subjects, although nearly all showed improvement. For receptive vocabulary, z scores ranged from –2.06 to 1.07. For 9 of the 10 subjects, scores were higher than at the original testing, with 7 subjects scoring within 1 SD of the mean currently. The increases ranged from z = 0.12 to 1.07. Deficits continue to be more substantial for expressive vocabulary. z scores ranged from –4.00 to 1.73, with 5 subjects scoring within 1 SD of the mean. For 3 subjects, the scores were measurably higher than at the original testing. Improvement in the scores of 3 other subjects in the current testing is assumed because of their inability to participate in the task at the original testing.
Receptive language z scores ranged from –2.22 to 1.00, with 6 subjects scoring within 1 SD of the mean. Increases in 2 subjects were measurable, whereas increase was assumed for 2 subjects who had not obtained any scores at the original testing. Expressive-language z scores ranged from –2.11 to 0.00, with 5 of the 9 verbal subjects scoring within 1 SD of the mean. Measurable differences could be calculated for only 2 subjects (both being increases); the others had not yielded scores on a structured instrument at the first test session because of low levels of functioning.
Notable increases in the quantity and complexity of verbal expressive language were noted in all the subjects. For most of the subjects, sentence length in number of words was judged to be grossly in the normal range for their ages, but the same could not be concluded regarding the morphologic and syntactic complexity of their sentences. More than 50% of the subjects continued to exhibit immature forms of tense marking and produced sentences incomplete in other morphologic markers. However, at least 30% of the subjects used conjoined and subordinate sentence structures, as well as sentences expressing a variety of functions (eg, questions, requests, negation). All parents reported increased development in expressive language. All subjects were still receiving speech-language intervention or had been receiving it within 1 year of the current testing.
Speech Quality
Overall, single-sound articulation accuracy had increased and intelligibility of connected speech in known contexts had increased, with overall intelligibility at the second test session ranging from 50% to 80% of utterances. The parents' judgments of their child's speech intelligibility was uniformly much higher than the assessor's. All parents reported increases in this area. For most of the subjects who exhibited noticeable misarticulations, these were in individual sounds, suggesting delayed development. For 1 subject, misarticulations were suggestive of speech dyspraxia, and 2 subjects exhibited mild slurring and imprecision.
Interval Medical Treatment
All subjects had received corticosteroids or ACTH before the first evaluation, and most subjects also received other immune-modulating or -suppressing medication. Of the 19 subjects, 14 continued to receive some form of treatment during at least part of the interval between evaluations. There was no influence of whether treatment was being administered in the interval between evaluations on change in test scores (see Table 2).
|
globulin, azathioprine, cyclophosphamide, pheresis, and, in 1 subject, an autologous stem cell transplant. In 5 children, no symptomatic relapses occurred with gradual withdrawal of medication despite severe opsoclonus-ataxia symptoms requiring substantial doses of corticosteroids or ACTH early in the course. Each had required 1 to 2 years of treatment, and each had been off of all treatment for at least 1 year at the time of the second evaluation. Three were on medication at the time of the first evaluation and tapered off before the second. Two subjects had been off of medication before the first evaluation. The clinical information regarding these 5 subjects is summarized in Table 3.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We found that a minority of children with opsoclonus-ataxia had an excellent outcome: cognitive, academic, behavioral, and motor function within the normal range and off of all immunotherapy. In addition, 4 of the 5 nonrelapsing subjects had no emotional or behavioral problems on second evaluation. One subject had some mild somatic and social problems. In our cohort, it was notable that all of the children with an excellent outcome had a course marked by lack of the usual relapses with tapering off of therapy or intercurrent illness despite severe initial symptoms and the need for aggressive early treatment to control opsoclonus-ataxia symptoms.
We could find no specific factors either in presentation or early treatment that differentiated patients who ultimately had frequent relapses and persistent developmental and neurologic deficits from the smaller group of children who had a monophasic course. Specifically, 4 of the 5 nonrelapsing subjects had very severe initial symptoms, whereas 1 had mild symptoms. Two subjects received primary therapy with ACTH, and 3 received oral corticosteroids. None received chemotherapy. It seems from these limited data that the difference between subjects with a monophasic course compared with those with a chronic, relapsing course was primarily a biological one that was not determined by treatment.
Findings indicate that the group as a whole made developmental progress from baseline to follow-up examination. Overall, cognitive scores increased in the group as a whole. Adaptive skills improved, but only slightly. The prominent behavioral and emotional problems seen on first evaluation did not seem to increase as subjects got older; in fact, we found a decrease in attention problems. Despite the variability in performance on speech and language measures, all the returning subjects had made noticeable gains in communication, some significantly so, progressing from minimal or no word production to verbal expression through adequately formed sentences.
It would be of great interest to prospectively follow immunologic parameters from early in the course of opsoclonus-ataxia to determine if there are differing patterns of immune activation that differentiate the children with a monophasic course from those with a chronic, relapsing course.
Even for children with opsoclonus-ataxia doing relatively well on treatment, concerns remain about long-term outcome, both the potential for late relapses and the possibility that subtle damage will continue to impair future learning and cognitive growth. The findings from our follow-up study highlight the need for early and periodic assessment of cognitive, motor, and behavioral functioning.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Addres correspondence to Wendy G. Mitchell, MD, Neurology Division, Childrens Hospital Los Angeles, 4650 Sunset Blvd, Box 82, Los Angeles, CA 90027. E-mail: wmitchell{at}chla.usc.edu
This work was presented in part at the 33rd Annual Meeting of the Child Neurology Society; October 14–16, 2004; Ottawa, Ontario, Canada.
No conflict of interest declared.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2005 American Academy of Pediatrics. All rights reserved. |
||
|
||
