30 April 2013. As the International Prodromal Research Network (IPRN) meeting moved into the afternoon of 21 April 2013 in Orlando, Florida, Robert Heinssen of the National Institute of Mental Health, Bethesda, Maryland, led a discussion on how to go about staging the prodrome. As with other illnesses (e.g., cancer), the treatment prescribed depends on the stage of illness, and a similar approach has been proposed for prodromal illness (see SRF related news story). Given the 11 promising studies on treating early stages of psychosis (Stafford et al., 2013), Heinssen suggested a framework that treated people according to the severity of their symptoms; for example, those with mild symptoms would receive some relatively benign treatment, whereas those with severe symptoms would receive something more aggressive, such as antipsychotics.
Though defining these stages would require careful precision, researchers were not convinced that more severe symptoms actually identified someone on the brink of psychosis. “Do we really know this is progressive?” asked Tyrone Cannon, noting that people do not necessarily move in one continuous direction from mild to worse prodromal symptoms. “The trajectories are all over the place,” he said. He preferred defining specific “states” of prodromal symptoms, with the understanding that there was not an obligatory progression through them. In a similar vein, Matcheri Keshavan suggested that prodromal symptoms could be stratified according to their biology. Because there may be different pathways to similar symptoms, Keshavan wondered whether it made sense to treat someone with, say, high inflammatory markers the same way as someone with high stress responses, even though they show the same symptoms.
Regardless of how the prodrome is eventually divvied up, the time seems ripe for education. Diane Perkins of the University of North Carolina, Chapel Hill, said that in general, clinicians on the front lines are unaware of the difference between frank psychosis and its sub-threshold signs. Rachel Loewy of the University of California, San Francisco, echoed this, saying that these practitioners have access to antipsychotic drugs, rather than cognitive behavioral therapy, and that it is unclear how best to bring the new treatments to their toolkits.
Finally, some suggested moving beyond psychosis to define a wider risk state that would include people showing early signs of depression or other mental illnesses, something Pat McGorry, who was not at the meeting, has advocated. Still, Heinssen said that currently there is more traction in identifying people on their way to psychosis than other conditions.
Whither the prodrome in DSM-5.1?
Scott Woods of Yale University, New Haven, Connecticut, spoke next on behalf of Alison Yung, who was unable to attend the meeting. Woods presented new data that could assuage some of the concerns raised about including attenuated psychosis syndrome (APS) in the DSM-5. One had to do with overdiagnosis of APS, which has been suggested in a study finding 20 percent of adolescents reported symptoms similar to APS (Kelleher et al., 2012). By contrast, Woods reported a lower rate of APS diagnosis of 2-8 percent in adolescents when a structured interview was used to assess symptoms. A second worry was that a diagnosis of APS would encourage overuse of antipsychotics. Data from the North American Prodrome Longitudinal Study (NAPLS) cohorts had originally suggested an increase in antipsychotic usage over time, but Woods showed that this increase evaporated when the data were adjusted for whether sites allowed APDs at the time of enrollment and other demographic and diagnostic variables. In fact, including APS in the next version of the DSM-5 could potentially decrease antipsychotic use, Woods argued, because people would not be diagnosed as psychotic. As for whether clinicians could reliably diagnose APS, Woods suggested that something more than the 30-minute education module used in the DSM-5 field trials might be necessary (see SRF related news story). Finally, regarding stigma surrounding an APS diagnosis, Woods said, “We need more data and less worry,” and suggested asking patients and clinicians.
Skeptical that a prodrome could be adequately described for the heterogeneous constellation of symptoms comprising a schizophrenia diagnosis, Robin Murray of the Institute of Psychiatry in London asked, “How can you have a risk syndrome for a fuzzy concept?” Murray noted studies finding that childhood psychotic experiences don’t necessarily foreshadow schizophrenia (Poulton et al., 2000), and can even increase risk for other disorders (Fisher et al., 2013). He also reported that efforts to find prodromal people miss many: though a clinic in South London identifies people at risk for psychosis (Fusar-Poli et al., 2012), most first-episode patients do not arrive via this clinic. Murray favored a broader approach, making clinics for youths with a wide range of psychiatric problems rather than focusing narrowly on psychosis risk. He even went so far as to suggest that psychosis was driven by anxiety and depression symptoms, and that these were the root problem. Several in the audience disagreed with this, saying that it could go both ways, with anxiety and depression resulting from psychosis, as well.
Will Carpenter of the University of Maryland began by thanking the audience for their help in developing the APS category for the DSM-5, saying, “You made a tedious life of DSM-5 become interesting.” He then outlined issues regarding APS that needed to be addressed for the next revision of the DSM-5. He noted that the DSM-5 field trials did not identify enough APS cases to make an estimate of how reliably it can be diagnosed, which is different from saying that APS is not reliably diagnosed. Carpenter said understanding how APS relates to other disorders may help, though ways to discern comorbid versus associated symptoms would be required.
As for the idea that a prodromal diagnosis was unnecessary because there was no treatment for it, Carpenter reminded the audience that clinicians themselves provide treatment. He suggested that this might, in fact, dilute effects found for treatments of APS, as the 11 completed trials used control groups receiving treatment as usual, rather than non-treated controls. Acknowledging the heterogeneous causes likely at work in the prodrome, Carpenter said that this shouldn’t hold the group back: treatments targeting secondary features such as a person’s resilience or compensatory mechanisms could work effectively and might be easier to modulate than trying to fix what originally went wrong. Finally, though he agreed that in practice psychiatrists treat symptoms rather than discrete categories of disorders, he maintained that an APS category was necessary in the DSM-5 manual to influence clinical practice, and to be noticed by regulatory, pharmaceutical, and other therapeutic entities.—Michele Solis.