Discussion
Our study represents the largest case series from India and has important observations. It illustrates the rarity of the syndrome in our population, similar to other studies. According to data from the United Kingdom, the estimated incidence of this syndrome was 0.18 new cases per million total populations per year.1 The prolonged duration of symptoms prior to referral and initial misdiagnoses highlights the fact that most pediatricians in our country are not aware of the syndrome. Pediatricians making the diagnosis of acute cerebellar ataxia in a toddler should be aware of possibility of opsoclonus-myoclonus syndrome. The existence of opsoclonus excludes the diagnosis of acute cerebellar ataxia. The presence of excessive irritability or sleep disturbances in coexistence of acute-onset ataxia, even in the absence of opsoclonus, should raise the suspicion of opsoclonus-myoclonus syndrome. Akin to previously published studies, we observed a similar median age at clinical presentation, no gender bias, common antecedent illness, acute/subacute onset of presentation, association of neuroblastoma, and response to immunomodulator therapy.1,6 However, in contrast with the reported high sensitivity (up to 95%) of metaiodobenzylguanidine scintigraphy scan for detection of neuroblastoma, we found that its sensitivity for tumor detection was low as compared to an abdominal CT scan. We also observed poor sensitivity of urine catecholamines in our cases. Similarly, Brunklaus and colleagues7 have reported poorer sensitivities for metaiodobenzylguanidine scintigraphy scan and urine catecholamines, reflecting the stage of maturation and predominant neurotransmitters of the cells comprising these tumors.8
In our series, only 4 of 11 cases had paraneoplastic opsoclonus-myoclonus syndrome. Previously published studies have reported a paraneoplastic etiology in about 50% of children.6 However, none of the investigative modalities has a 100% sensitivity to detect neuroblastoma; also, this tumor is known to have spontaneous regression. Therefore, a possibility of a very small tumor that escaped detection even after investigations and regressed spontaneously cannot be excluded.9
We observed a good therapeutic response with immunomodulators, including adrenocorticotrophic hormone, intravenous immunoglobulin, and parenteral and oral corticosteroids. It is difficult to compare these therapies because of the small number of patients. One child with paraneoplastic opsoclonus-myoclonus failed to respond with adrenocorticotrophic hormone but had a significant response with intravenous immunoglobulin. In recent studies, rituximab, an anti-CD 20 monoclonal antibody, has been shown to be effective and safe for treatment of pediatric opsoclonus-myoclonus syndrome.10,11 We were not able to use rituximab in our patients because of cost constraints. Recently, Tate and colleagues12 conducted the largest study on comparison of adrenocorticotrophic hormone-based immunotherapies for pediatric opsoclonus-myoclonus syndrome. They reported greater efficacy of adrenocorticotrophic hormone–based multimodal therapy compared with adrenocorticotrophic hormone alone and greater response to adrenocorticotrophic hormone than corticosteroid-based therapy.12 There are also anecdotal reports of benefit with topiramate, clonazepam, and oral high-dose dexamethasone pulses.13⇓–15
Neurologic relapses in our series were less than those reported by Tate and colleagues (up to 52 %).6 This could perhaps be due to the small number of patients in our study. Also, in our series, the outcome was better in those with idiopathic opsoclonus-myoclonus syndrome whereas in the study by Tate and colleagues, the outcome was independent of etiology. Interestingly, Bataller and colleagues16 have also observed better outcome of idiopathic opsoclonus-myoclonus syndrome in adult patients. Mitchell and colleagues17 described development and neurologic sequelae in 17 children with opsoclonus-ataxia caused by neuroblastoma. They observed substantial developmental sequelae and increased deficits in older children. They postulated that this represents a progressive encephalopathy rather than a time-limited single insult.17 We did not observe progressive developmental and behavioral syndrome in our patients, which could be due to small number of patients of paraneoplastic etiology, short-follow-up, and lack of objective psychometric testing. Brunklaus and colleagues18 studied clinical predictors of long-term outcome and demonstrated that severe initial symptoms and very young age at disease onset are important risk factors of developing long-term sequelae.
To conclude, opsoclonus-myoclonus syndrome is a rare but one of the few immunomodulator-responsive pediatric neurologic disorders. Screening for an occult neuroblastoma is necessary in children with the syndrome. Paraneoplastic opsoclonus had a poor outcome in our experience