Saccade inhibition in schizophrenia : towards phenotyping based on brain function

Abstract

Schizophrenia is a complex psychiatric illness which is caused by multiple genes and gene-environment interactions. This complex etiology has a negative impact on the power of molecular genetic studies. To improve this situation, attempts are made to define traits that are more intimately related to the genetic factors of schizophrenia, and have a more straightforward genotype-phenotype relationship. These traits are called endophenotypes. The deficit of patients to suppress reflexive saccades during antisaccades has been suggested as endophenotype for schizophrenia. Recent evidence suggests that functional magnetic resonance imaging (fMRI)-brain activation maps can further strengthen the genotype-phenotype relationship. In a series of studies, the suitability of brain activation maps during antisaccades as endophenotype for schizophrenia is assessed. In the first study, an event related fMRI design was developed for measuring the brain activation associated with saccades and saccadic inhibition. This study demonstrated that fMRI could pick up brain activation in all areas that are known to be involved in oculomotor functioning from studies in primates. In addition, the results of this study stress the importance of using event related fMRI for this particular saccade paradigm. An endophenotype should be related to the illness. In the second study, the fMRI design was used to find the functional brain abnormalities that underlie the saccadic inhibition deficit in schizophrenia. It was found that patients did not activate the striatum during inhibition of saccades. These results indicate that the impairment of patients in saccadic inhibition can be related to a dysfunction in the striatum, or in the frontostriatal circuitry. An endophenotype is more frequently present in those with genetic risk for schizophrenia. Accordingly, the third study found the same striatal abnormality in healthy siblings of patients during antisaccades. However, the siblings did not demonstrate an behavioural abnormality, suggesting that fMRI activation maps can increase sensitivity. When subjects are phenotyped based on brain activation, age differences between subjects can ad noise to the data. In the fourth study, effects of aging on brain activation during the antisaccade paradigm were assessed in a group covering a broad age range. Age did not only alter the cognitive strategy for the task, but also the neuronal vascular coupling. Good endophenotypes should be stable traits. The fifth study assessed the reproducibility of brain activation during antisaccades over a one week period in a group of healthy subjects. The results demonstrated large differences between subjects in the test-retest reliability of brain activation maps, which was caused largely by stable differences between subjects in signal to noise ratios. Differences between subjects in signal to noise can result in overestimation of between subject variance. In conclusion, fMRI is capable of detecting saccade related brain activation and abnormalities in patients and their siblings, even when no behavioural abnormalities could be detected. Hence the use of fMRI could potentially strengthen the genotype-phenotype relationship. Although fMRI is not yet capable of to obtain reliable activation maps in all subjects, this situation is likely to improve in the future with increasing MR-field strengths and more sophisticated acquisition techniques.