CSF Cerebrospinal fluid IMD Inherited metabolic disorderAIM For the 9% to 16% of children with cerebral palsy (CP) who have normal brain imaging, further testing for metabolic and ⁄ or genetic conditions has been recommended. This study aimed to identify a cohort of children with CP with normal magnetic resonance imaging (MRI), clinically review and describe the cases, and assess the value of testing for inherited metabolic disorders in these children.METHOD Children with congenital CP born from 1999 to 2005 were selected from a population register. Normal MRI reports were identified and the scans reassessed. Children whose scans were performed before 18 months were excluded, as were children with spastic CP (Gross Motor Function Classification System [GMFCS] level I). The remainder were reviewed clinically and offered investigations. RESULTSOf 730 children identified, 515 had available imaging and 54 were confirmed as normal.Cases with non-spastic CP and those with milder clinical severity were more likely to have normal imaging. Twenty-three children (17 males, six females; mean age 6y 11mo, SD 1y 10mo, range 3y 0mo to 10y 0mo) were reviewed clinically and offered investigations. Twelve children had spasticity (11 with diplegia, one quadriplegia), three had dyskinesia, five ataxia, and three hypotonia. Two children functioned in GMFCS level I, 11 in level II, seven in level III and three in level IV. Four children with spasticity had unusual features. No alternative diagnoses were made.INTERPRETATION Although important to consider in individual cases, comprehensive metabolic testing failed to clarify the aetiology of CP further in this large cohort of children with normal MRIs, even those with atypical features.Cerebral palsy (CP) is a clinical description, not an aetiological diagnosis. It describes a group of disorders of the development of movement and posture, causing activity limitation, that are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain.1 CP may be classified by the predominant type of motor disturbance as spastic, dyskinetic, or ataxic, and sometimes hypotonic.In children with CP, magnetic resonance imaging (MRI) can be a useful tool to help establish a cause or to elucidate the likely pathophysiology. Imaging findings have been described and classified in cohorts of children with particular types or topographical patterns of motor impairment and the results have been summarized in three systematic reviews.2-4 There have only been three published studies that were population based and included all subtypes of congenital CP; two of these are previous work of our research group. Evidence from these sources indicates that, although most children with CP have demonstrable brain abnormalities on MRI, an estimated 9% to 16% have normal scans. [2][3][4][5][6][7] In two of the population studies, an association between normal MRI and mild functional severity was found, 5,6 leading us to hypothesize that MRI may have failed to pick up some of the more subtle...

This letter is linked to the letter by Whitehouse. on pages 226-232.SIR-We thank Dr Whitehouse for his interest in our paper on metabolic testing in children with cerebral palsy (CP) and normal brain imaging, 1 and for his constructive comments on the interpretation and implications of the results of our study. His suggested methods for providing an upper confidence limit in the circumstance where there were no positive results were most helpful.Our study was certainly limited by its small sample size, and was probably not originally designed to answer the study question definitively. Rather, our project had the objective of highlighting an under-researched area that has important clinical implications. We had hoped that positive results would help clinicians decide when to engage in the process of testing, but in the end our negative results left the issue open for debate.An important part of our interest in doing the study related to the process itself. This involved reviewing the status of all eligible children and contemplating a practical process for deciding which children should be investigated further, and what testing might be performed. In the end we could not ethically justify performing costly, time-consuming, or invasive procedures in children exhibiting a typical pattern of spasticity of a mild severity, or in those with an already established causal pathway. If we had included all children with CP and normal imaging, our sample size would have been larger, and presuming we obtained the same diagnostic yield, the upper confidence interval would have been lower.Clearly our negative findings should be treated with caution. We definitely concur with Dr Whitehouse in his assertion that children with abnormal imaging suggestive of an inherited metabolic disorder should undergo metabolic testing, as should those with other features consistent with inherited metabolic disorders. These children are not eligible for inclusion on the Victorian Cerebral Palsy Register, and were not included in our study. On the other hand, the Victorian Cerebral Palsy Register includes some children with hypotonic CP, a motor type that is not accepted under the CP banner in many parts of the world. Three children with hypotonic CP underwent further testing for inherited metabolic disorders as part of our study. We also agree with Dr Whitehouse that the likely diagnostic yield of further investigation of children with magnetic resonance imaging in the setting of non-progressive motor impairment is still unclear. We suggest that larger, possibly multicentre, studies would provide better power to answer the study question and, hopefully, produce stronger evidence to assist clinicians in the management of these diagnostically difficult cases.