NEW YORK – A team of Irish scientists is developing a prognostic test for psychotic disorders, with the hope that such a tool could help stratify patients presenting with clinical symptoms, as well as be used in tandem with therapeutic interventions.
The researchers at the Royal College of Surgeons Ireland recently received €1.3 million ($1.6 million) from Wellcome Trust's psychosis innovation flagship to support the development of the blood-based test. The team, led by David Cotter, a professor of molecular psychiatry at RCSI in Dublin, relied on mass spectrometry as a discovery tool to narrow in on 10 proteins of interest that might predict which patients presenting with clinical symptoms might progress to full-blown psychosis years later.
According to Cotter, there is great interest in developing a test that can catch psychotic disorders before the patients become essentially invalids, depending wholly on care providers for in some cases decades, making them expensive to both care for and treat.
"Researchers have moved from studying schizophrenia to early psychosis, putting all their money in the basket to predict who will develop a psychotic illness," said Cotter. "We can already identify those at risk; that is easy enough because it is based on clinical symptoms," he said. "But only about 20 percent of them will transition to a psychotic disorder."
Yet if psychiatrists were able to predict who will transition, they can be treated with preventative drugs, perhaps slowing or mitigating the onset of their diseases. "That is what our biomarkers do," said Cotter. "And if we can confirm this, it will transform the way we deliver services."
To develop the list of proteins, Cotter and colleagues looked at blood samples obtained from a prospective cohort of 344 individuals at high clinical risk of psychosis and followed them for several years to observe who did and didn't progress to a psychotic disorder. They analyzed the samples using a Q Exactive mass spectrometer made by Thermo Fisher Scientific. They used machine learning tools to identify proteins of interest in the early blood samples to predict those who made the transition.
The work, discussed in an August 2020 paper in JAMA Psychiatry, found that the panel of proteins could predict those who transitioned to a psychotic disorder in 93 percent of high-risk cases, as well as those did not transition in 80 percent of cases.
With the new grant, RCSI's team will work with partners at Cardiff University in Wales, the University of North Carolina at Chapel Hill, the University of Bristol, Trinity College Dublin, King's College London, Yale University, the University of Melbourne, and University College Dublin's Conway Institute to further assess the use of the proteins as a prognostic test.
"We want to replicate our findings and develop the test," said Cotter. He noted that the new study will not be a prospective cohort study, and instead will rely on samples and clinical data collected by various partners. "These samples already exist," said Cotter. "We will have the samples sent to us and we will analyze them."
Cotter noted that an unnamed commercial partner is also working with the investigators, and that ideally, the markers could be transitioned to another format, such as an enzyme-linked immunosorbent assay, and made available as a test kit for high risk patients. The utility of such a test, however, would be incumbent on whether it could be linked to therapeutic interventions, and shown to make a difference in outcome. "We will have to change the treatment and see if the outcomes of those patients are better in the real world," said Cotter.
He credited mass spectrometry as a discovery tool in the study. "We would have not been able to do this without mass spectrometry-driven proteomics discovery," said Cotter. "It is powerful, fast, and reliable," he said. "Technology has driven these findings and the predictions are real."
David Mongan, a PhD student in Cotter's group, agreed. "Mass spectrometry is great from a discovery point of view, but to be used in a regular clinic, any test has to stand up to robust reproducibility," Mongan said. He speculated the panel of proteins could be made available as an ELISA, or perhaps on another mass spectrometry-based platform.
Regardless of how the test reaches the market and when – it could take five to 10 years, Mongan acknowledges – the need is "huge," he said. "I have worked in a psychosis prevention service and it is very difficult to predict who is most at risk of developing a psychotic disorder," he said. He added that such a test could also be integrated into other diagnostic and therapeutic studies.
"The clinical implications could be massive, because you can tell people they are at particular risk," said Mongan. "But obviously it needs to be tied to some kind of intervention ideally to actually prevent it in the first place."
Oliver Schubert, an associate professor of psychiatry at the University of Adelaide in Australia is familiar with the RCSI team's work but not affiliated with the study. He said the work is of "great interest to the field" as it shows that a protein-based blood test in childhood could predict who will likely develop significant mental health problems later on.
According to Schubert, previous studies using various types of clinical and biological data have made similar attempts to predict which patients will transition but have generally concluded that biological information on its own is insufficient to make solid future predictions.
"In fact, clinical factors, such as related to a patient's upbringing, their symptom patterns, and their family history have been found superior to biological tests in terms of predictive accuracy," said Schubert. He said the RCSI study "for the first time shows that a protein blood test, irrespective of other data points, could be sufficient to highlight which children are most at risk."
However, much more work needs to be done around these proteins before they find their way into routine clinical use. Schubert pointed out that the proteins assessed are part of immune and coagulation pathways. While the community has long been aware of the link between the immune system and psychiatric disorders, the changes in the levels of these proteins merits further inquiry.
"How exactly these changes of blood protein levels work to bring about mental illness years later will need to be subject to further study," said Schubert. He said the changes might be potentially linked to other, yet unidentified underlying pathologies. "It is also not yet clear how these changes develop over time, and whether they can be influenced in some way, such as through medications or other therapeutic interventions to reduce illness risk," he said.
Until these questions are answered, the implementation of the test in clinical practice would be "challenging," said Schubert. "For example, what advice would one give to someone with a positive test result, in the absence of definitive counter-measures?" he asked. "Where things currently stand, that advice would be quite generic, such as 'Don't ever take drugs, look after yourself, exercise regularly, avoid too much stress.'"
And while a positive test result in childhood might lead to faster detection of mental illness in adulthood, and thus to faster access to treatment, it might also embroil some difficult ethical issues around the quality of life of at-risk children and their families, Schubert added.
"In all, this study is a milestone towards better definition and detection of ill mental health," he said. "It opens up exciting avenues towards truly preventive psychiatric practice."