Goals

Psychotic disorders, including schizophrenia and affective psychotic disorders are common, affecting about 5% of the population.  Development of these disorders usually results in significant life-long disability, and psychotic disorders rank among the top 10 causes of disability worldwide.  The ability to predict these outcomes prospectively using objective criteria is crucial to the development of preventive approaches.  Also crucial to this mission is increased knowledge of the neurobiological and genetic mechanisms underlying conversion to psychosis among at risk youth.

NAPLS represents a collaboration of 8 clinical research centers based at Emory University, Harvard University, University of North Carolina, University of Calgary, University of California Los Angeles (UCLA), University of California San Diego (UCSD), Yale University, and Zucker Hillside Hospital.  In a prior phase of our consortium, we pooled clinical and psychosocial data obtained on a large sample of patients (N=291) who had been ascertained using a standardized set of operational diagnostic criteria for a ‘prodromal’ risk syndrome.  Risk for onset of psychosis in this population was 35% after 2 and ½ years of follow-up, with a decelerating rate of conversion over this period.  Moreover, prediction algorithms incorporating baseline clinical and psychosocial variables dramatically improved positive predictive power (~80%) compared with the prodromal criteria alone, but achieved only modest sensitivity (~40%).  In the current phase of NAPLS our goals are to improve upon these prediction algorithms by incorporating biological measures and to test the possible differential course of change in biological indicators in those who convert to psychosis.

The potential utility of biological assays in elucidating predictors and mechanisms of psychosis in the prodromal population is thus far based exclusively on analyses of data collected at individual sites with small samples.  The UCLA team, working collaboratively with investigators in a prodromal program based in Melbourne, Australia, has recently demonstrated a significantly steeper rate of gray matter reduction in prefrontal cortical regions in prodromal patients who convert to psychosis compared with those who do not over a 1-year follow-up period.  This pattern of accelerated change in prefrontal regions is mirrored in a sample of first-episode schizophrenia patients compared with age- and gender-matched healthy controls over a 2-year follow-up period.  Together, these data suggest that during the prodromal and early phases of schizophrenia, there is an exaggeration of the regressive neuromaturational processes (programmed cell death, synaptic pruning) normative to late adolescence and early adulthood, changes that may participate in the pathophysiology of psychosis onset.  To test this model rigorously, while at the same time accounting for the marked heterogeneity in outcomes among prodromal and early psychosis patients, will require sample sizes many times larger than those available in any single site.  In addition, ideally, any investigation into the course of gray matter reduction in the prodromal phase of psychosis would incorporate information from other assessment modalities (genomics, proteomics) that can reveal molecular mechanisms for the steeper rate of change in the converting group.

In the current phase of the NAPLS consortium, we are conducting a prospective, longitudinal study of 720 prodromal patients and 240 matched healthy controls incorporating neuroimaging, electrophysiological, hormonal and genomics assessments. In terms of psychosis prediction, we seek to determine whether biological abnormalities preceding psychosis onset contribute to prediction of psychosis independently from that of the best performing clinical algorithms and whether they can be combined with the clinical measures to enhance predictive utility.  Based on analyses of small subsets of subjects incorporating selected biological measures at individual sites, we anticipate that the addition of biological measures to the algorithms will result in substantially improved sensitivity (~80%), while retaining high positive predictive power (~80%).  In addition, we anticipate heterogeneity in optimal risk prediction profiles among different subgroups of psychosis cases, and, as an additional extension of our prior work, we will determine whether these profiles differentially predict different DSM-IV diagnostic outcomes of schizophrenic vs. affective vs. other psychoses.

This study also holds great promise for elucidating the course of change in neurobiological indicators of vulnerability to schizophrenia in relation to the onset and course of psychosis.  Although it is well documented that the late adolescent/early adult period is critical for the emergence of psychotic symptoms, there have been no large-scale, longitudinal studies focusing on biological indicators during this period.  The psychosis prodrome provides a unique window on the unfolding pathophysiology of illness, without the clouding effects of disease chronicity and long-term treatment that plague studies of patients with established illness.  By employing a multi-site collaborative framework, we will achieve a prospective, longitudinal dataset about 20 times larger than in any prior study.  This substantial increase in statistical power will enable detection of small to moderate differences, both cross-sectionally and longitudinally, between those who do and do not convert to psychosis and healthy controls, while accounting for the marked heterogeneity in risk characteristics and outcomes inherent to the prodrome population.  This study will enable us to determine whether prodromal patients who convert to psychosis show a steeper rate of change in neurobiological risk indicators compared to non-converters and healthy controls and to isolate the brain systems involved (e.g., dorsolateral prefrontal cortex, superior temporal gyrus, hippocampus) across multiple levels of analysis (anatomical, physiological, behavioral).  If a steeper rate of change occurs with psychosis onset, the findings would suggest deviant adolescent brain maturational processes as playing a role and encourage search for molecular mechanisms underlying these changes.  We will be in a unique position to test particular genomic (DNA, RNA), hormonal (cortisol), and environmental (perinatal, drug abuse, stressful life events, viral) factors that may contribute to the observed changes in anatomy, physiology, and behavior over time.