Visual event-related potentials in cirrhotic patients without overt encephalopathy

Ambulant patients with cirrhosis and no clinical evidence of encephalopathy were screened for impaired brain function by neuroelectrophysiological testing dependent on cognitive function. Infrequent large checkerboard visual stimuli were randomly interleaved with frequent small ones to elicit P300 event-related potentials (ERPs). Three ERP components, N200, P3a and P3b, were derived from the electroencephalogram (EEG) by computer averaging. The use of 10% contrast and a minimum of four precisely placed scalp electrodes were found to be necessary for optimal separation of ERPs from sensory evoked potentials. Visual ERPs, onset/offset and pattern-reversal visual evoked potentials (VEPs), the spontaneous EEG and the time taken to complete a standard number connection test (NCT) were obtained from 20 normal adult subjects and 19 age-matched patients with histologically-confirmed cirrhosis and no clinical evidence encephalopathy. The latencies and amplitudes of evoked potentials and the alpha rhythm of the EEG were determined. In 6 of the 19 patients the latencies of P3a and/or P3b exceeded the corresponding mean for controls + 2 standard deviations of that mean. In 4 other patients the NCT was prolonged. In all of the patients the N200, VEPs and alpha rhythm of the EEG were normal.

(i) Optimal isolation of ERPs is critically dependent on stimulus contrast and electrode placement; (ii) ERPs appear to be more sensitive than primary sensory evoked potentials or the EEG in detecting impaired brain neuroelectrophysiological function; and (iii) Cirrhotic patients without overt encephalopathy in whom P3a and/or P3b latencies are prolonged may have subclinical hepatic encephalopathy.