It seems that in neuropathology the dictum of real estate applies, it is all about location, location, location. In essence location of pathology constrains the possible number of causative agents and forecasts the natural history of a particular disorder. So in autism the ongoing debate is whether abnormalities are located within the cerebral cortex or in subcortical structures like the limbic system or cerebellum.
For many Neurologists the clinical symptoms of autism would certainly suggest some type of widespread abnormality of the cerebral cortex. The cerebral cortex is the seat for higher cognitive functions and abnormalities of the same could easily provide for many of the deficits of the condition including socialization, language, restricted interests, sensory hypersensitivities, and seizures. On the other hand the main clinical features of cerebellar disorders (e.g., incoordination, imbalance and problems in stabilizing eye movements) are seldom, if ever, seen in autistic individuals. In fact, during my clinical training examining symptoms in patients with cerebellar disorders disclosed remarkable findings that could not be easily confused. Although the cerebellum is not necessary for the initiation of movement, once started, they became quite erratic in size and direction. Thus the overall deficits are quite striking.
It may be said that the so-called cerebellar findings in autism are not as blatantly overt as they are developmental in nature. An analogy may be drawn to young patients with cerebellar tumors who receive an ablation of this structure early in their life. These patients may not develop overt coordination abnormalities, but they also do not develop autism.
Very interestingly the cerebellum has significant projections to the frontal lobes that implicates this structure in higher cognitive functions. While the midline portion of the cerebellum (called the vermis) regulates the accuracy of trunk, head and eye movements, the lateral cerebellar hemispheres are engaged when a person tries to solve a difficult problem. It is said that even the generation of appropriate verbs (and therefore language) may be modulated by the activity of the lateral cerebellar hemispheres.
It is of interest that Bauman and Kemper’s original case report showed cell loss in both cerebellar hemispheres, site of the cerebellum’s higher cognitive functions. More to the point, the pathology was quite specific in pinpointing a selective vulnerability for some big neurons called Purkinje cells. Ever since their original report, Bauman and Kemper added an additional nine cases to their series, all with similar findings. Contrary to neuroimaging studies showing abnormalities in the midline cerebellum (vermis) this brain region never showed any histological abnormality. Bauman and Kemper suggested that the lack of concomitant gliosis and absence of changes within the olivary nucleus suggested that Purkinje cell loss occurred early during brain development.
A good number of more recent studies have corroborated the presence of Purkinje cell loss in autism, but the nature of the change remains controversial. Bauman and Kemper never quantitated any neuronal changes within the olivary nucleus of autistic individuals. Their opinion was based solely on subjective appraisal of their microscopic slides. This was criticized by Bailey and others who believed that subjective reports were inaccurate in lieu of both the large variations in neuronal density within this nucleus and the large distances between examined cells. Similarly, the use of Nissl stains to screen for the possible presence of gliosis stands against standard knowledge in neuropathology. Contrary to the opinion of Bauman and Kemper, studies using immunocytochemistry for GFAP have shown the presence of gliosis along the Purkinje cell layer of the cerebellum. The reported gliosis does not involve tissue reorganization nor glial scar formation. The presence of GFAP positive astrocytes is not a response to a focal lesion but rather a reactive process to Purkinje cell loss.
A significant portion of autistic individuals suffer from epilepsy, take anticonvulsants, or engage in self-injurious behaviors (such as head banging). Purkinje cell loss may thus be the result of excitatory glutaminergic damage (seizures), toxicity (phenytoin), or contusive injury (head banging). A developmental pathology seems a far-flung possibility under these conditions. In fact, the normal development of the cerebellar folds (called folias) in autism argues against the large number of Purkinje cell loss having happened during brain development. It is unsurprising that Bauman and Kemper have reassessed their original opinion. In a study comparing immunocytochemistry and Nissl staining they indicated the possibility that agonal conditions and postmortem artifacts could have accounted for inadequate staining and the lower Purkinje cell counts reported in their original studies (Whitnery et al., 2008).
It is interesting that Purkinje cell loss appears to be a reactive process (as shown by GFAP staining) that happens during the postnatal life of autistic individuals. The ability to quickly identify the same has lead to its use as a marker when developing animal models for the condition. Similarly, its presence in autism has been regarded as a core pathology of the condition regardless of whether the same is a phenomenon secondary to seizures, medication or trauma. It is time to reassess our research priorities. Purkinje cell loss is a real phenomenon in autism, but the same is not developmental in nature. It appears to be a reactive phenomenon to multiple comorbidities.
Whitney, E. R., Kemper, T. L., Bauman, M. L., Rosene, D. L., & Blatt, G. J. (2008). Cerebellar Purkinje cells are reduced in a subpopulation of autistic brains: A stereological experiment using calbindin-D28K. Cerebellum, 7, 406–416.