Documente Academic
Documente Profesional
Documente Cultură
Original article
Article history: Objective: To compare the clinical data at diagnosis, treatment and neurological outcome in
Received 16 June 2009 34 children with opsoclonus–myoclonus syndrome (OMS) associated with a detected
Received in revised form neuroblastoma or not.
7 December 2009 Study design: This is a multicentric retrospective study of 34 children presenting with OMS
Accepted 22 December 2009 from four pediatric centers diagnosed between 1988 and 2008.
Results: Twenty-two patients had OMS associated with a neuroblastoma. These patients all
Keywords: had neuroblastomas with favourable prognostic features; all underwent surgery, six
Opsoclonus–myoclonus syndrome received chemotherapy. Twelve children had OMS without a detected neuroblastoma. For
Dancing eye syndrome OMS, the main treatment in all children was corticotherapy (n ¼ 33), but immunoglobulins
Neuroblastoma (n ¼ 13), cyclophosphamide (n ¼ 4) and rituximab (n ¼ 4) were also given. In the 27 OMS
Developmental delay patients with or without neuroblastoma whose follow up was greater than two years, the
neurological outcome was evaluated: 59.3% had neurological sequelae, including motor,
praxic and/or language sequelae (n ¼ 9), persistent ataxia (n ¼ 6) and moderate motor deficit
(n ¼ 3). No significant difference in neurological outcome was noted between the two
patient groups.
Conclusion: Our retrospective study provides further evidence that OMS with or without
a detected neuroblastoma is the same disease, whose major challenges are the
* Corresponding author.
E-mail address: gudrun.schleiermacher@curie.net (G. Schleiermacher).
1090-3798/$ – see front matter ª 2009 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ejpn.2009.12.005
european journal of paediatric neurology 14 (2010) 400–409 401
it seems possible that a small NB escaping detection or one catecholamine metabolites, and MYCN amplification), as well
which had already regressed may have occurred, suggesting as oncological treatment and outcome were recorded.
that OMS with or without a detected NB may be the same
entity. The mechanism leading to this condition is probably 2.2. Neurological data
autoimmune but the physiopathology remains poorly under-
stood.10–14 OMS is seen in 2–3% of children with NB10,15 and a NB For patients with OMS both with or without NB, the degree of
is found in 50–80% of children with OMS.2 Children with NB ataxia, opsoclonus, myoclonus and sleep or mood disturbance
presenting with this paraneoplastic syndrome have an excel- were evaluated at diagnosis, during treatment and during
lent oncological outcome. Most often, the oncological treat- follow up, and the scoring system was assessed a posteriori
ment consists of surgery only, and sometimes chemotherapy. based on the patients’ records. The grading scale used is rep-
However, the neurological outcome of pediatric patients resented in Fig. 1, adapted from that developed by Mitchell and
with OMS is more compromised: neurological sequelae have Pike.2 Infectious and immunological investigations were per-
been reported in 70–80% of patients, including motor, formed particularly when no NB was identified: viral serologies
language, praxic and cognitive deficits.2,13,15,16 Delayed iso- (adenovirus, enterovirus, coxsackie, EBV, HHV6, VZV, CMV,
lated cerebellar atrophy has been described on MRI during HSV, HTLV1 and 2, human Parvovirus), bacterial serologies
long-term follow up.17 In order to improve neurological (Mycoplasma pneumoniae, Chlamydia pneumoniae, Borrelia burg-
outcome, different treatment modalities for OMS including dorferi, syphilis), but were not performed in patients with an
steroids, cyclophosphamide, intravenous immunoglobulins identified NB. Study of the CSF consisted of cells count, protein
and more recently rituximab have been used, with many side electrophoresis, interferon dosage, as well as chromatography
effects and a variable efficacy.18–23 But to date, the treatment of of cerebral amino acids in some cases. Brain imaging (MRI or
OMS is not standardized. Thus, the hallmark of this syndrome tomodensitometry) was performed at diagnosis and two years
is on the one hand the frequent relapse of neurological symp- later for some of the children. Lastly, at the moment of the last
toms when immunosuppressive treatment is decreased, and neurological follow up, the neurological status was recorded.
on the other hand the threat of long-term cognitive sequelae.
The goal of this study was to compare the clinical presen- 2.3. Statistical analysis
tation, treatment and neurological outcome of OMS associ-
ated with NB versus OMS without NB. For comparison of continuous variables between the two
patients group, the Mann–Whitney test (MW) was used. For
comparison of categorical, the c2 test was used. The level of
2. Patients and methods significance for all statistical tests was p < 0.05. Statistical
analysis was performed using the medcalc software.
This retrospective study comprised pediatric patients who
were treated for OMS with or without a NB, from two large
pediatric oncology centers (Institut Curie and Institut Gustave 3. Results
Roussy), and two large pediatric neurology centers (Necker
hospital and Kremlin Bicêtre hospital), between 1988 and Clinical data of the whole population are summarized in
2008, in the Ile-de-France (Paris, France) region, with more Table 1.
referrals to these specialised centers occurring within more
recent years. The 34 children (12 males, 22 females) were
followed in a multidisciplinary setting. Eligible patients were Table 1 – Oncological and neurological characteristics of
those who had a diagnosis of OMS defined by the presence of the whole population (n [ 34).
at least three of the following criteria: opsoclonus, ataxia/ Whole NBþ NB#
myoclonus, behavior changes and NB. The medical records of population (n ¼ 22) (n ¼ 12)
these patients admitted for OMS were reviewed. All children (n ¼ 34)
were screened for a primary NB tumor by various imaging
Percentage of detected NB 64.7% – –
methods, including abdominal sonography, neck, chest and Median age (months) 20.8 18.1 26.2
abdomino-pelvic CT-scan or MRI and I123 metaiodo-benzyl- Median neurological score at 9.3 10.1 7.3
guanidine (MIBG) scintigraphy, as well as with urinary cate- diagnosis
cholamine metabolites detection. The detection methods
Treatment
were heterogeneous due to the large recruitement period and Corticosteroids alone 11 5 6
evolution of clinical practice. Corticosteroids and second 20 17 3
line treatment
2.1. Oncological data Unknown 3 – 3
Neurological outcome
For each child with OMS and NB, oncological data at diagnosis Recovery 38.2% 36.4% 41.7%
and during the follow up were collected. At diagnosis, the age (n ¼ 13) (n ¼ 8) (n ¼ 5)
of onset of symptoms, the symptoms that lead to the diag- Sequelae 52.9% 63.6% 33.3%
(n ¼ 18) (n ¼ 14) (n ¼ 4)
nosis of the tumor (OMS or not), the characteristics of the
Unknown 8.8% - 25%
tumor (histological features, fixation on MIBG scintigraphy,
(n ¼ 3) (n ¼ 3)
stage according to INSS classification, detection of urinary
european journal of paediatric neurology 14 (2010) 400–409 403
Abbreviations: M: male, F: female, U: unknown, CR: complete remission, RT: residual tumor.
3.1. OMS associated with NB follow up (Table 3). At diagnosis, all except one had severe
stance and gait abnormalities (score >2). All patients had
3.1.1. Oncological data opsoclonus except one, 14 patients presented with mood
In 22 patients, an OMS associated with a NB was observed. The disturbance and all but one had ataxia. The neurological score
oncological data for these 22 patients are listed in Table 2. at diagnosis ranged from 2 to 15 (median 10.9).
Their age at diagnosis ranged from 9 to 33 months (median: Patient 11, 16 and 22 underwent a cerebral MRI after the
18.1 months). Follow up ranged from 12 months to 16 years onset of OMS (range 18 months to 2 years), which was normal.
(median 72 months) (Table 3). OMS lead to diagnosis of NB in all
patients except one, in whom NB was discovered by abdominal 3.1.3. Treatment and follow up of OMS
pain, and clinical signs of OMS appeared secondarily. The NB All patients received corticotherapy during first line treatment
of these children had mainly features of good prognosis (Table except one who recovered spontaneously just after the
2). Surgical resection was the only oncological treatment in 16 surgical resection and two who received immunoglobulins
patients, and six others received chemotherapy associated (Table 3). Oral corticotherapy was the main treatment: pred-
with surgery (neoadjuvant and/or adjuvant). nisone (posology: 1–2 mg/kg/day) in 11 cases, either continu-
ously (n ¼ 10) or intermittently (n ¼ 1, patient 2), and
3.1.2. Neurological data hydrocortisone (10–14 mg/kg/day) in 6 cases. Three patients
The degree of ataxia, myoclonus, opsoclonus and mood or received IV methylprednisolone pulses and three received
sleep disturbance were recorded at diagnosis and during the dexamethasone pulses. The median duration of
404
Table 3 – Neurological presentation, treatment and outcome on the patients with OMS and detected neuroblastoma.
Patient Neurologic initial presentation Chemotherapy Initial ttt Other ttt Administration Length of Follow Neurological
modalities of corticotherapy up outcome
Stance Gait Arm and Opsoclonus Mood/ Total/ corticotherapy
hand Behaviour 15
function
developmental
outcome
up. Ten patients received immunoglobulins (Ig): in two cases
Ig were given first without any success (posology: 1 gram per
7 years Recovery
Chemotherapy: VPC for VP16 and carboplatine, CADO for cyclophosphamide, doxorubicin and vincristine, PE for platinum and etoposide, *, neoajuvant, #, adjuvant, 2: number of cures.
kilo per day, during 2 days), whereas 8 other patients received
delay
Ig because OMS was corticoresistant or corticodependant.
Three patients received cyclophosphamide, and four had rit-
Follow
4 years
8 years
course of rituximab. Only two patients (NB#) were treated for
relapse of OMS after complete recovery with steroids as first
line treatment, when steroids were totally stopped. 14
patients (9 NBþ and 5 NB#) were corticodependant, i.e. the
Administration
corticotherapy
Continuous oral
Continuous oral
Abbreviations: ttt: treatment, Ig: immunoglobulin/HC: hydrocortisone/MP: methylprednisolone/DXM: dexamethasone/RTX: rituximab/U: unknown.
relapses were observed when decreasing the corticotherapy.
HC MP pulses
Corticosteroid
2 VPC-CADO*
(median 7.3). All patients had severe gait and stance distur-
bance (score >2), and only one patient had no opsoclonus. All
these patients received a corticotherapy as first treatment
Total/
15
8
10
1
2
2
2
a cerebral MRI was performed five years after the onset of the
1
1
0
2
3
2
3
21
22
Table 5 – Treatment and outcome of the patients with OMS, associated or not with neuroblastoma.
Treatment Neurological outcome
statistically significant, there was a trend towards an older age the sensitivity of all imaging methods increased during the
at diagnosis in the population of patients with OMS without study period, small NB tumors could have been missed in the
NB (Mann–Whitney test, p ¼ 0.07). oldest cases of this series. More recently, standardized imaging
strategies have been proposed to optimize neuroblastoma
3.3.2. Neurological status at diagnosis diagnosis: MRI is the best modality to diagnose a primary tumor
The NB-positive patients compared to the NB-negative OMS and evaluate its local extension, except for abdominal tumors
patients had a statistically significant higher neurological for which CT remains the best modality. Metastatic extension
score at diagnosis: median 10.1 (range 2–15), versus median should be assessed with MIBG scan and liver sonography.25,26
7.3 (range 5–13), respectively (MW, p ¼ 0.01). In a next step, the Neurological features were reported according to a scaling
score of each symptom at diagnosis was compared between system, based a posteriori on the clinical description in the
the two groups. There was a significant difference between medical files. In this study, patients with NB-positive OMS had
the two groups considering stance (MW, p ¼ 0.03), gait (MW, a higher neurological score at diagnosis. Due to the retro-
p ¼ 0.03) and the arm and hand function (MW, p ¼ 0.05). All spective analysis of non standardized data, the results may be
these symptoms’ scores were significantly higher in the NB- difficult to interpret This underlines the importance of
positive patients. There was no significant difference in terms a homogeneous scoring system and a multidisciplinary
of opsoclonus (MW, p ¼ 0.17) and mood and behavioral approach in all future prospective studies.
changes (MW, p ¼ 0.35). Neurological treatment for OMS was heterogeneous,
especially because of the large period of this retrospective
3.4. Neurological outcome study. Thirty-three in 34 patients received corticotherapy. The
different therapeutic modalities did not seem to have an
In Table 5, the neurological outcome of all the patients (with or impact on outcome in this small series. However, it would be
without a NB) whose follow up was greater than two years is interesting to compare the efficacy of these heterogeneous
shown. Among 27 patients, 11 recovered completely (40.7%). modalities, in order to standardize the corticotherapy, and to
Sixteen (59.3%) had neurological deficit: nine had motor, praxic specify the indications of a second line treatment. A long-term
and/or language sequelae, and six had persistent ataxia or corticosteroid administration seems to be the most wide-
myoclonus. There is no significant difference in the neurolog- spread practice, even if it involves many side effects and an
ical outcome (presence of sequelae vs. recovery) between the unpredictable efficacy.27 Other immunomodulatory therapies,
two groups of patients (with and without NB, c2 test p ¼ 0.27). like intravenous immunoglobulins, cyclophosphamide or rit-
The neurological outcome was not different between the groups uximab were tried, but it is difficult to determinate their effi-
of children who received oral or intravenous corticotherapy (c2 cacy, as they often were tried after corticosteroids for
test, p ¼ 0.14), and between the groups of children who received corticoresistant or corticodependant OMS, but also because of
chemotherapy for treatment of the NB or not (c2 test, p ¼ 0.68). the few patients who received them. These other treatments
were always given in addition to corticosteroids in our study,
and not as first line treatment, except immunoglogulins, given
4. Discussion as first line treatment in two patients who received cortico-
steroids as second line treatment. A 29 patients’ retrospective
This study is a retrospective analysis of the neurological analysis had found a significative association between
features of 34 patients with OMS. a treatment with chemotherapy and a good neurological
In 22 cases, OMS was associated with a NB of favourable outcome.28 This observation had not been confirmed by other
prognosis. Among 12 NB-negative OMS patients, in only 4 studies1,15 and our study do not provide any conclusion
could an infectious etiology be suspected. However, despite regarding the efficiency of the chemotherapy on the neuro-
multiple investigations having been performed, the search for logical outcome because of the very few patients who received
an infectious etiology was not standardized. Furthermore, as it (only six).
408 european journal of paediatric neurology 14 (2010) 400–409
Regarding neurological outcome, only one previous study Long-term survival. Study of the French Society of Pediatric
has already compared the neurological outcome of patients Oncologists]. Arch Pediatr 2000;7(6):621–8.
with OMS with or without NB (8 with NB, 3 without), and the 2. Matthay KK, Blaes F, Hero B, Plantaz D, De Alarcon P,
Mitchell WG, et al. Opsoclonus myoclonus syndrome in
authors concluded that OMS without NB had a better neuro-
neuroblastoma a report from a workshop on the dancing eyes
logical outcome.24 These results were not confirmed in our syndrome at the advances in neuroblastoma meeting in
series, as among our patients, the frequency of neurological Genoa, Italy, 2004. Cancer Lett 2005;228(1–2):275–82.
sequelae was similar between OMS patients with and without 3. Pohl KR, Pritchard J, Wilson J. Neurological sequelae of the
NB. dancing eye syndrome. Eur J Pediatr 1996;155(3):237–44.
In order to evaluate the prevalence of neurological 4. Dassan P, Clarke C, Sharp DJ. A case of poststreptococcal
sequelae in children with OMS and NB, we compared our opsoclonus–myoclonus syndrome. Mov Disord 2007;
22(10):1490–1.
neurological results with the neurological data of the main
5. Peter L, Jung J, Tilikete C, Ryvlin P, Mauguiere F. Opsoclonus–
retrospective studies on the subject. In Russo’s series,28 20/29 myoclonus as a manifestation of Lyme disease. J Neurol
children had persistent neurologic sequelae (speech delay, Neurosurg Psychiatr 2006;77(9):1090–1.
cognitive deficits, motor delay, behavioral disturbances). The 6. Chemli J, Ketata S, Dalhoumi A, Ajmi H, Hassayoun S, Fodha I,
nine children who recovered were not different from the et al. [Opsoclonus–myoclonus syndrome associated with
others of the group in term of age at diagnosis and duration of Mycoplasma pneumoniae infection.]. Arch Pediatr 2007;
14(8):1003–6.
symptoms. But 6/9 children with complete recovery received
7. Zaganas I, Prinianakis G, Xirouchaki N, Mavridis M.
chemotherapy. In Plantaz’s series,1 10/16 had persistent
Opsoclonus–myoclonus syndrome associated with
neurological deficits, and there was no relation between cytomegalovirus encephalitis. Neurology 2007;68(19):1636.
neurological outcome, age at diagnosis, duration of neuro- 8. Medrano V, Royo-Villanova C, Flores-Ruiz JJ, Sempere AP,
logical symptoms or oncological treatment, especially Mola-Caballero de Roda S. [Parainfectious opsoclonus–
chemotherapy. In Rudnick’s study,15 14/19 children developed myoclonus syndrome secondary to varicella-zoster virus
neurological or behavioral sequelae. In Hayward’s study,17 9/ infection]. Rev Neurol 2005;41(8):507–8.
9. van Toorn R, Rabie H, Warwick JM. Opsoclonus–myoclonus in
11 children had mild to severe neurological sequelae. Brain
an HIV-infected child on antiretroviral therapy – possible
MRI was performed in five children and showed cerebellar immune reconstitution inflammatory syndrome. Eur J
atrophy in each of them. In Mitchell’s series,29 all of the 17 Paediatr Neurol 2005;9(6):423–6.
patients had neurological deficit. Pohl3 studied 54 patients 10. Pranzatelli MR. The neurobiology of the
with OMS: leukemia was found in one case, NB only in three opsoclonus–myoclonus syndrome. Clin Neuropharmacol 1992;
cases. 91% had neurological deficit. Taken together, 70/99 15(3):186–228.
11. Pranzatelli MR, Tate ED, Wheeler A, Bass N, Gold AP,
patients (70.7%) had at long-term mild to severe neurological
Griebel ML, et al. Screening for autoantibodies in children
deficit. In our patients whose follow up was greater than 2
with opsoclonus–myoclonus–ataxia. Pediatr Neurol 2002;
years, 16/27 (59.3%) had a persistent mild to severe neurolog- 27(5):384–7.
ical deficit, confirming the poor neurological outcome previ- 12. Pranzatelli MR, Travelstead AL, Tate ED, Allison TJ,
ously described in the litterature. Moticka EJ, Franz DN, et al. B- and T-cell markers in
In conclusion, in NB-positive and NB-negative OMS, opsoclonus–myoclonus syndrome: immunophenotyping of
although neurological features at the time of diagnosis seem CSF lymphocytes. Neurology 2004;62(9):1526–32.
13. Raffaghello L, Conte M, De Grandis E, Pistoia V.
to be slightly worse in the NB-positive group, neurological
Immunological mechanisms in opsoclonus–myoclonus
outcome is equally poor, justifying a homogeneous treatment
associated neuroblastoma. Eur J Paediatr Neurol; 2008.
strategy. These results further support the hypothesis that 14. Antunes NL, Khakoo Y, Matthay KK, Seeger RC, Stram DO,
OMS with or without NB could be the same disease, a NB Gerstner E, et al. Antineuronal antibodies in patients with
having escaped detection in the latter group. Because of the neuroblastoma and paraneoplastic opsoclonus–myoclonus. J
rarity of this disease, a multidisciplinary and international Pediatr Hematol Oncol 2000;22(4):315–20.
collaboration to improve the knowledge about OMS is 15. Rudnick E, Khakoo Y, Antunes NL, Seeger RC, Brodeur GM,
Shimada H, et al. Opsoclonus–myoclonus–ataxia syndrome in
required, in order to set up a standardized diagnostic proce-
neuroblastoma: clinical outcome and antineuronal
dures and treatment, and to work towards improvement of antibodies-a report from the Children’s Cancer Group Study.
the neurological outcome. An international trial protocol in Med Pediatr Oncol 2001;36(6):612–22.
order to evaluate prospectively the outcome of patients with 16. Koh PS, Raffensperger JG, Berry S, Larsen MB, Johnstone HS,
OMS would be interesting. This protocol should include brain Chou P, et al. Long-term outcome in children with
MRI to assess cerebellar abnormalities, videos and standard- opsoclonus–myoclonus and ataxia and coincident
ized neuropsychological tests, performed by the same person neuroblastoma. J Pediatr 1994;125(5 Pt 1):712–6.
17. Hayward K, Jeremy RJ, Jenkins S, Barkovich AJ, Gultekin SH,
to make the interpretation easier.
Kramer J, et al. Long-term neurobehavioral outcomes in
children with neuroblastoma and opsoclonus–myoclonus–
ataxia syndrome: relationship to MRI findings and anti-
neuronal antibodies. J Pediatr 2001;139(4):552–9.
references 18. Pranzatelli MR, Tate ED, Travelstead AL, Longee D.
Immunologic and clinical responses to rituximab in a child
with opsoclonus–myoclonus syndrome. Pediatrics 2005;
1. Plantaz D, Michon J, Valteau-Couanet D, Coze C, 115(1):e115–e119.
Chastagner P, Bergeron C, et al. [Opsoclonus–myoclonus 19. Pranzatelli MR, Tate ED, Travelstead AL, Barbosa J,
syndrome associated with non-metastatic neuroblastoma. Bergamini RA, Civitello L, et al. Rituximab (anti-CD20)
european journal of paediatric neurology 14 (2010) 400–409 409
adjunctive therapy for opsoclonus–myoclonus syndrome. J 25. Swart JF, de Kraker J, van der Lely N.
Pediatr Hematol Oncol 2006;28(9):585–93. Metaiodobenzylguanidine total-body scintigraphy required
20. Rostasy K, Wilken B, Baumann M, Muller-Deile K, Bieber I, for revealing occult neuroblastoma in opsoclonus–myoclonus
Gartner J, et al. High dose pulsatile dexamethasone therapy in syndrome. Eur J Pediatr 2002;161(5):255–8.
children with opsoclonus–myoclonus syndrome. 26. Brisse H, Edeline V, Michon J, Couanet D, Zucker J,
Neuropediatrics 2006;37(5):291–5. Neuenschwander S. [Current strategy for the imaging of
21. Bell J, Moran C, Blatt J. Response to rituximab in a child with neuroblastoma]. J Radiol 2001;82(4):447–54.
neuroblastoma and opsoclonus–myoclonus. Pediatr Blood 27. Ertle F, Behnisch W, Al Mulla NA, Bessisso M, Rating D,
Cancer; 2006. Mechtersheimer G, et al. Treatment of neuroblastoma-related
22. Burke MJ, Cohn SL. Rituximab for treatment of opsoclonus– opsoclonus–myoclonus–ataxia syndrome with high-dose
myoclonus syndrome in neuroblastoma. Pediatr Blood Cancer; dexamethasone pulses. Pediatr Blood Cancer 2008;50(3):683–7.
2006. 28. Russo C, Cohn SL, Petruzzi MJ, de Alarcon PA. Long-term
23. Wilken B, Baumann M, Bien CG, Hero B, Rostasy K, neurologic outcome in children with opsoclonus–myoclonus
Hanefeld F. Chronic relapsing opsoclonus–myoclonus associated with neuroblastoma: a report from the Pediatric
syndrome: combination of cyclophosphamide and Oncology Group. Med Pediatr Oncol 1997;28(4):284–8.
dexamethasone pulses. Eur J Paediatr Neurol; 2007. 29. Mitchell WG, Davalos-Gonzalez Y, Brumm VL, Aller SK,
24. Hammer MS, Larsen MB, Stack CV. Outcome of children with Burger E, Turkel SB, et al. Opsoclonus–ataxia caused by
opsoclonus–myoclonus regardless of etiology. Pediatr Neurol 1995; childhood neuroblastoma: developmental and neurologic
13(1):21–4. sequelae. Pediatrics 2002;109(1):86–98.