Autismo

Brain (1998), 121, 889–905

A clinicopathological study of autism
A. Bailey,1 P. Luthert,2,* A. Dean,2 B. Harding,3 I. Janota,2 M. Montgomery,2,* M. Rutter1 and P. Lantos2
Child Psychiatry Unit, and 2Department of Neuropathology, The Institute of Psychiatry and 3Department of Histopathology, Institute of Child Health, London, UK
1MRC

Correspondence to: A. Bailey, MRC Child PsychiatryUnit, The Institute of Psychiatry, De Crespigny Park, Denmark Hill, London SE5 8AF, UK *Present address: The Department of Pathology, The Institute of Ophthalmology, University College, London, UK

Summary
A neuropathological study of autism was established and brain tissue examined from six mentally handicapped subjects with autism. Clinical and educational records were obtained and standardizeddiagnostic interviews conducted with the parents of cases not seen before death. Four of the six brains were megalencephalic, and areas of cortical abnormality were identified in four cases. There were also developmental abnormalities of the brainstem, particularly of the inferior olives. Purkinje cell number was reduced in all the adult cases, and this reduction was sometimes accompanied bygliosis. The findings do not support previous claims of localized neurodevelopmental abnormalities. They do point to the likely involvement of the cerebral cortex in autism.

Keywords: autism; neuropathology; megalencephaly; cortical dysgenesis Abbreviations: ADI Autism Diagnostic Interview; GFAP glial fibrillary acidic protein

Introduction
Autism is a severe developmental disorder characterized byimpairments in reciprocal social interaction and communication, restricted and stereotyped patterns of behaviour and interests, and an onset before 3 years of age (World Health Organization, 1992). The core disorder affects approximately four in 10 000 children, and is much commoner in males, in a ratio of ~4 : 1. The syndrome was first described by Leo Kanner (1943); he assumed that affectedchildren were of normal intelligence and for several decades the disorder was thought to be psychogenic. An organic basis was first suggested by the finding that three-quarters of sufferers are mentally handicapped (Lockyer and Rutter, 1969) and that at least one-quarter develop epilepsy (Rutter, 1970; Gillberg and Steffenburg, 1987). Subsequently autism was often assumed to be an unusual consequence ofbrain damage, caused either by medical disorders or obstetric hazards. More recently it has been appreciated that only a minority of cases of autism are associated with medical causes of mental handicap, then most commonly with tuberous sclerosis or Fragile X (Rutter et al., 1994). The findings from twin and family studies suggest that the vast majority of idiopathic cases arise on the basis ofstrong specific genetic influences © Oxford University Press 1998 (Bailey et al., 1996). Thus, autism usually appears to represent a severe expression of a specific disease process. Many regions of the brain have been implicated in the genesis of autism, but the neurobiological basis of the disorder remains unknown. Autism is a rare disorder which was described relatively recently and there have beenonly a few post-mortem studies. Darby (1976) reviewed 33 diverse cases and found no consistent abnormalities. Two of the four cases reported by Williams et al. (1980) had associated disorders (phenylketonuria and probable Rett’s syndrome); in one of the two idiopathic cases pyramidal cell dendritic spine density was reduced in the mid-frontal gyrus and the number of cerebellar Purkinje cells was alsodiminished. Coleman et al. (1985) undertook cell counts in several cortical regions from a single case and two control subjects. Consistent differences were not found, although the glia : neuron ratio was smaller in the autistic brain than in the two control subjects. Recent examination of the brainstem of this case (Rodier et al., 1996) revealed a hypoplastic facial nucleus and superior olive....