Acquired (Non-Wilsonian) hepatocerebral degeneration with "shunt myelopathy"
It is well known that patients with severe liver disease may suffer from a reversible neurologic condition known as hepatic encephalopathy [1,6]. Some of these patients suffer from a more permanent neurologic disorder, particularly if they have surgically induced or spontaneous porto-systemic shunts and experience episodes of severe hepatic encephalopathy with coma. Victor, Adams, et al. reported the features of AHCD in detail in a series of 27 cases . These symptoms and signs included progressive dementia, dysarthria, involuntary movements (including tremor, asterixis, and choreoathetosis), ataxia of limb and gait, and mild pyramidal tract signs, typically in a patient with severe hepatic disease and recurrent episodes of coma. Less commonly reported is a progressive spastic myelopathy in association with AHCD [4,12]. Only about 15 cases have been reported in the world literature with detailed pathological descriptions [2,3,7,9,13].
The present case report has several distinctive features. Whereas most reported cases resulted from a surgically created porto-systemic shunt as a treatment for cirrhosis, a few other cases, including this one, are thought to result from spontaneously created shunts (e.g., varices) with the same functional consequences. The published cases usually had Alzheimer II cells in the brain, but only a few had the severe spongy degeneration in the deep cortex illustrated here. Recently, hepatocerebral degeneration has occasionally been documented in vivo by MRI [8,9]. Few electron microscopic studies of human Alzheimer II cells and hepatocerebral degeneration have been published [5,11]. The present report suggests that the vacuolation results from the disruption of astrocytes.
The cause of the myelopathy and the reason why only some patients with severe liver disease develop this complication remains uncertain. In some of the reported cases , the myelopathy was discontinuous and the disease process appeared to be mainly demyelinative, in contrast to the Wallerian degeneration in the corticospinal tracts observed in our patient. Because the hepatocerebral degeneration involved the motor cortex, but the tract degeneration was present only in the spinal cord and was not visible at or above the level of the medulla, the possibility of a dying back phenomenon, as proposed by Pant et al.  must also be considered in the present case. Although no obvious loss of Betz cells was observed, there may well have been sufficient interference with the normal function of these cells, possibly from disrupted astrocyte function, to give rise to a distal degeneration.
Acknowledgments: We thank Ruth Grajcer and Roxana Norville for excellent technical assistance. This work was supported by the VA Medical Research Program.
Contributed by David C. Spencer and Lysia S. Forno