Predicting the Future from First Psychotic Episodes
12 September 2006. From a first psychotic episode, equal numbers of patients improve or get worse. Finding a way to predict the path patients will follow is an area of intense research, with the ultimate goal of allowing physicians to offer the best possible treatment as early as necessary to individual patients.
A new study from Stephen Wood and colleagues at the University of Melbourne in Australia offer evidence that measurements of brain chemistry at first psychotic episode, using magnetic resonance spectroscopy (MRS), can predict disease outcome. The study, which appears in the September issue of the Archives of General Psychiatry, shows that lower n-acetylaspartate/creatine ratio (a marker of neuronal dysfunction) in the prefrontal cortex at the first occurrence of psychosis correlated with a poorer clinical picture 18 months later.
Another factor tied to poor outcomes in the short term is duration of untreated psychosis. A second new paper looks at the important question of whether that association holds up with longer follow-up. The study, from Eadbhard O’Callaghan and colleagues at University College Dublin, Ireland, shows that indeed, as long as four years later, a longer duration of untreated psychosis correlates with poorer functioning and more symptoms in patients. The work, which is published in the September issue of the British Journal of Psychiatry, strengthens the case for early intervention as a way to improve treatment outcomes.
The MRS study by Wood and colleagues builds upon previous work showing that reduction in n-acetylaspartic acid (NAA) in the prefrontal cortex in chronic schizophrenia is associated with poorer cognition, longer duration of illness, and greater negative symptoms. NAA is considered a marker of neuronal health, and is reduced in brain regions under conditions of neuronal loss. Reductions of NAA might also signal lower neuronal metabolism or problems with myelination, both of which could contribute to the prefrontal cortex abnormalities reported in schizophrenia.
However, studies from Wood and coworkers show that this marker, normally given as the n-acetylaspartate/creatine (NAA/Cr) ratio, is quite variable in early disease (Wood et al., 2003), and they wondered if this variation might presage later disease outcomes. To get at that question, the researchers measured NAA/Cr in two brain areas (left prefrontal cortex and left mediotemporal lobe) in 46 patients with first-episode psychosis who were treated at their center in Melbourne. The subjects were followed up with outcome measures after 18 months of comprehensive treatment. Outcomes were defined as scores on three standard evaluations: the Global Assessment of Functioning Scale (GAF), Clinical Global Impression Scale (CGI) and the Social and Occupational Functional Assessment scale (SOFAS). In addition, the researchers measured the number of in-patient admissions in the follow-up period.
A low NAA/Cr ratio in the left prefrontal cortex was significantly related to worse outcomes in all four measures. Even when adjusted for negative symptoms at baseline and premorbid GAF scores, which have been related to poorer outcomes, the NAA/Cr measure was still significantly associated with three of the four outcome measures, and was nearly significant for the fourth. The prefrontal cortex NAA/Cr ratio independently explained between 17 and 30 percent of the variance in the four different outcome measures, far more than age, duration of untreated psychosis, or baseline negative symptoms. Putting the NAA/Cr ratio together with premorbid GAF scores plus three other predictive factors (baseline negative symptoms, age at imaging, and duration of untreated psychosis) explained about half the variance in outcomes at 18 months.
In contrast to these results, NAA/Cr ratio in the mediotemporal lobe did not correlate with outcome. This supports the idea that changes in that region occur later in the disease progression.
These results suggest that impaired prefrontal function early in the course of schizophrenia may define a subset of patients with poor prognosis and poor response to treatment. The promise is that combining MRS with the other markers of negative symptoms, premorbid functioning, DUP, and age could help identify patients who would benefit from more intensive treatment early on. More studies will be needed, however, to find out if the association holds beyond the 18-month mark and persists over longer courses of disease.
Persistence is a critical question for all prognostic markers for long-term conditions like schizophrenia, and one that the Dublin group addresses in their study. To determine if the duration of untreated psychosis, which has been linked to short-term outcomes, is also tied to long-term outcome, first author Mary Clarke and colleagues followed up on a group of 166 patients 4 years after first presentation. The mean duration of the untreated psychosis was 17.9 months (s.d. = 32.1, range 0.25–240), with a median of 5 months. The mean prodrome, available for 152 patients, was 21.3 months (s.d. = 35.7, range 0–177), with a median of 3 months. At both study entry and follow-up, the researchers assessed psychopathology with the Positive and Negative Syndrome Scale (PANSS) and global functioning with the GAF scale.
Those patients with a longer duration of untreated psychosis had a reduced likelihood of remission, displayed poorer function, and increasing psychopathology, allowing the authors to conclude that, “The association between duration of untreated psychosis and functional outcome supports the likely benefits of early intervention in psychosis.”—Pat McCaffrey.
Wood SJ, Berger GE, Lambert M, Conus P, Velakoulis D, Stuart GW, Desmond P, McGorry PD, Pantelis C. Prediction of functional outcome 18 months after a first psychotic episode: a proton magnetic resonance spectroscopy study. Arch Gen Psychiatry. 2006 Sep ;63(9):969-76. Abstract
Clarke M, Whitty P, Browne S, McTigue O, Kamali M, Gervin M, Kinsella A, Waddington JL, Larkin C, O'callaghan E. Untreated illness and outcome of psychosis.
Br J Psychiatry. 2006 Sep;189:235-40. Abstract