End of lines and boxes

In this issue, Joel publishes a thoughtful iteration of the circuit models that have dominated thinking about basal ganglia pathophysiology for almost 15 years. These models have generally been fruitful in the sense that they have led to new experiments, and somewhat surprisingly, provided support and guidance for novel therapeutic approaches. For example, these models all emphasized the importance of the subthalamic nucleus as a key regulator of basal ganglia output, and provided a good deal of impetus for the implementation of pallidotomy and subthalamic deep brain stimulation as treatments for parkinsonism. The relative success of these models is surprising in view of the fact that they are simple anatomic constructs with little effort to incorporate anything beyond the crudest notions of physiology. Many aspects of these models have been challenged appropriately. As Joel makes clear, the anatomy of the basal ganglia is much more complicated than the original versions of these models present. In particular, Joel takes pains to incorporate advances in knowledge of striatal output pathways that add new dimensions of complexity to basal ganglia circuitry. While the original models incorporated the idea of “closed” cortico-basal ganglionicthalamocortical loops and had simple definitions of the so-called direct and indirect pathways, Joel and others emphasize the interconnection between these circuits. This approach has the virtue, explicated rigorously by Joel, of allowing relatively discrete lesions to have consequences that ramify beyond the major function of the circuit primarily affected. In Joel’s construct, many phenomena associated with Huntington disease can be explained by pathology in the associative portions of the striatum. Joel also points out some predictions of the model that can be evaluated experimentally, a necessary feature of any good model. Joel’s model preserves one of the strongest and worst features of the original models. These models are not physiologically based, but are rather elaborate examples of clinico-pathologic correlation. This is a strong method for attempting to understand the basis of clinical phenomena and has real power in assisting the development of new therapies. This method is limited ultimately by how much physiological meaning we can extract from studying clinical phenomena. An instructive comparison is the study of perception. In terms of really understanding brain function, it seems that study of perceptual processes, particularly visual perception, is farther advanced than any other area of neuroscience. Perceptual neuroscientists benefit from substantial advantages. Unlike motor physiologists, sensory physiologists can correlate neural activity with precisely defined inputs. These inputs can be decomposed into well-understood physical properties, and the distinguished tradition of psychophysics provides useful measurement tools. These tools can be applied to human experiments, animal experiments, and more recently, functional imaging experiments. Very little work on basal ganglia function can claim the relative success of sensory physiology in giving mechanistic accounts of sensory, particularly visual, function. Models of basal ganglia pathophysiology are formulated in the absence of good understanding of normal basal ganglia function(s). Joel’s model pushes the clinical correlation method to its limits. She makes a number of associations between specific deficits in Huntington disease subjects and speculated dysfunctions within defined pathways. These associations, however, are based on constructs with relatively shaky foundations. She incorporates well, for ex*Correspondence to: Roger L. Albin, M.D., Neuroscience Laboratory Building, 1103 E. Huron, Ann Arbor, MI 48104-1687. E-mail: ralbin@umich.edu Received 27 November 2000; Accepted 29 November 2000 Published online 9 May 2001 Movement Disorders Vol. 16, No. 3, 2001, pp. 405–406 © 2001 Movement Disorder Society Published by Wiley-Liss, Inc.