Cortical Folding May Predict Antipsychotic Drug Response
August 28, 2013. Examining the folds of the brain's outer layer, the cortex, provides clues about which patients with first-episode psychosis will respond well to antipsychotics, reports a new study published online August 14, 2013, in JAMA Psychiatry. Lena Palaniyappan of the U.K.'s University of Nottingham and colleagues found that non-responders (those whose symptoms did not improve with subsequent antipsychotic treatment) showed less cortical gyrification in several brain regions than did treatment responders.
Although many patients with first-episode psychosis (FEP) experience symptom reduction and even remission following antipsychotic treatment, symptoms remain for some. Early treatment response is one of the best predictors of functional outcome (Emsley et al., 2007), yet determining just who will respond well remains a challenge. Several neuroimaging studies have probed the link between treatment response and brain structure in psychosis. For example, alterations in cortical gyrification are present after obstetric complications, a well-replicated risk factor for schizophrenia, and are also associated with treatment-resistant psychotic symptoms (Cachia et al., 2008).
In the current study, Palaniyappan and colleagues examined gyrification across the entire cortical surface using magnetic resonance imaging (MRI). FEP patients were scanned as soon as possible after their first clinical contact to allow for minimal antipsychotic drug exposure. The severity of psychotic symptoms was assessed on the day of the brain scan and again after 12 weeks of antipsychotic treatment, using the Positive and Negative Syndrome Scale (PANSS).
The researchers used a 3D reconstruction of the cortical surface to compute local measurements of gyrification at thousands of vertices across the surface—termed local gyrification indices—each indicative of the amount of cortex buried within a particular point. Palaniyappan and colleagues found decreased gyrification in 80 FEP patients compared to 46 controls in several cortical regions: the left middle/inferior frontal gyrus, precentral gyrus, and precuneus and the right middle frontal gyrus and inferior parietal region.
Reviewing response to treatment
When the researchers split the patients into two groups—those who responded to subsequent antipsychotic treatment (n = 40) and those who did not (n = 40)—they also found differences in gyrification. Compared to treatment responders, non-responders exhibited reduced cortical folding (or hypogyria) in the insula, superior frontal, and rostral middle frontal regions in the left hemisphere and the inferior and superior temporal cortex in the right hemisphere.
When compared to controls, both responders and non-responders displayed significant hypogyria of the left lingual gyrus. However, while the responders' gyrification looked similar to controls' in all other areas, non-responders had significantly less gyrification than controls did in a number of cortical regions, including the bilateral middle frontal gyrus, the right superior/inferior temporal cortex, angular gyrus, and medial occipital cortex, and the left posterior cingulate and precuneus.
Thus, "…already at illness onset, patients with FEP who subsequently do not respond to treatment have significant cortical folding defects compared with patients who subsequently respond and with healthy controls, while those who go on to respond are virtually indistinguishable from the controls," conclude the authors.
Although methodological differences preclude a direct comparison of the current surface-based approach with previous voxel-based morphometry studies, the authors also noted that their results are broadly consistent with earlier findings of reduced gray matter volume in non-responders (Szeszko et al., 2012).
Delving into diagnosis
As a secondary outcome measure, Palaniyappan and colleagues also measured gyrification as a function of diagnosis. They found that patients with nonaffective psychosis (schizophrenia, schizoaffective disorder, schizophreniform disorder, and psychosis not otherwise specified; n = 56) had reduced gyrification in several brain areas compared to those with affective psychosis (bipolar disorder and major depression with psychosis; n = 24). Because cortical folding develops most rapidly early in life, these results may be consistent with the idea that schizophrenia is characterized by greater neurodevelopmental abnormalities than bipolar disorder or depression.
The researchers found no significant interaction between diagnosis and treatment response. However, responders with nonaffective psychosis had reduced gyrification of the bilateral insula and left medial orbitofrontal, dorsolateral prefrontal, and superior temporal sulci relative to those with affective psychosis, while non-responders did not have a significant difference in gyrification between diagnostic groups. Despite some overlap, the gyrification deficits that were associated with poor treatment response were largely separate from those that characterized nonaffective psychoses.
"Our study provides crucial evidence of neuroimaging markers that can be used early in psychosis to predict prognosis in clinical settings," conclude the authors, adding that early identification of non-responders may lead to alternative, more effective treatment plans for these individuals.—Allison A. Curley.
Palaniyappan L, Marques TR, Taylor H, Handley R, Mondelli V, Bonaccorso S, Giordano A, McQueen G, Diforti M, Simmons A, David AS, Pariante CM, Murray RM, Dazzan P. Cortical Folding Defects as Markers of Poor Treatment Response in First-Episode Psychosis. JAMA Psychiatry. 2013 Aug 14. Abstract
Comments on News and Primary Papers
Comment by: Robert B. Zipursky
Submitted 6 September 2013
Posted 6 September 2013
Palaniyappan et al. demonstrated that subjects with a first episode of psychosis (FEP) had a reduction in the extent of cortical gyrification in multiple brain areas compared to healthy comparison subjects. Notably, non-responders showed more prominent hypogyria than responders did in a number of frontal and temporal areas irrespective of whether the underlying diagnosis was of an affective or non-affective psychosis.
Previous studies using structural MRI have established that subjects with FEP have smaller cerebral gray matter volumes (Zipursky et al., 1992) but have left open the question of whether these differences reflect the result of early neurodevelopmental processes that are aberrant versus progressive degenerative losses taking place more proximal to the onset of psychosis. Palaniyappan et al. suggest that measures of cortical gyrification, which is believed to be particularly active during intrauterine growth and early infancy, are more indicative of a developmental than a degenerative process. Their finding that subjects with FEP have reductions in cortical gyrification is interpreted as supporting the neurodevelopmental origin of these differences.
The debate over whether the differences in brain structure found in FEP are developmental or degenerative in nature (Zipursky et al., 2012) is of particular relevance to the question of treatment response. If the neuropathology of schizophrenia involves a progressive degenerative process, then intervening as early as possible to halt this process and bring about a remission of symptoms becomes of critical importance. There has, therefore, been intense interest in the possibility that the duration of untreated psychosis (DUP) is an important determinant of treatment response in FEP (Perkins et al., 2004). However, it has remained unclear whether the association reported between DUP and treatment response is a causal one; it is possible that individuals with longer DUPs might have more longstanding developmental problems that result in a more insidious onset and less robust response to treatment. This report by Palaniyappan et al. reinforces the view that poor treatment response is more likely to be understood as reflecting more severe longstanding developmental differences in brain structure. These possibilities need not be mutually exclusive; early developmental changes in the brain and later progressive changes occurring closer to illness onset might both contribute to poor treatment response. Poor clinical outcomes could also relate to DUP through a range of different mechanisms that do not require that a progressive process is taking place in the brains of those with an FEP (Zipursky et al., 1992).
Most patients with FEP will have a robust response to antipsychotic medication. We were able to demonstrate a number of years ago that FEP patients who did not respond to low doses of haloperidol had smaller cortical gray matter volumes than responders did (Zipursky et al., 1998) and that even with higher doses of haloperidol, these individuals had less improvement than the low-dose responders (Zhang-Wong et al., 1999).
The results of Palaniyappan et al. are in keeping with these findings. The focus of much research on treatment response has been to understand the determinants of antipsychotic response. As most FEP patients will have a robust response, it may be of greater interest to ask what factors underlie poor response. The 20-30 percent of individuals with FEP who do not have a remission of symptoms are likely to have a poor outcome in the longer run as well. Understanding that poor response may reflect the result of longstanding neurodevelopmental differences, as suggested by this study by Palaniyappan et al., may lead to more specialized approaches to identifying and treating poor responders earlier in their illness course (Agid et al., 2007). The development of more sophisticated, stratified approaches to the management of schizophrenia would be enhanced if neuroimaging markers could be proven to be valuable in predicting treatment response. The study by Palaniyappan is an important step in this direction. Further research will be required to ensure that medication effects are not confounding the measurement of cortical gyrification and that measures of gyrification are clinically meaningful determinants of treatment response.
Agid O, Remington G, Kapur S, Arenovich T, Zipursky RB. Early use of clozapine for poorly responding first-episode psychosis. J Clin Psychopharmacol. 2007;27(4):369-73. Abstract
Perkins D, Lieberman J, Gu H, Tohen M, McEvoy J, Green A, et al. Predictors of antipsychotic treatment response in patients with first-episode schizophrenia, schizoaffective and schizophreniform disorders. Br J Psychiatry. 2004;185:18-24. Abstract
Zhang-Wong J, Zipursky RB, Beiser M, Bean G. Optimal haloperidol dosage in first-episode psychosis. Can J Psychiatry. 1999;44(2):164-7. Abstract
Zipursky RB, Lim KO, Sullivan EV, Brown BW, Pfefferbaum A. Widespread cerebral gray matter volume deficits in schizophrenia. Arch Gen Psychiatry. 1992;49(3):195-205. Abstract
Zipursky RB, Zhang-Wong J, Lambe EK, Bean G, Beiser M. MRI correlates of treatment response in first episode psychosis. Schizophr Res. 1998;30(1):81-90. Abstract
Zipursky RB, Reilly TJ, Murray RM. The myth of schizophrenia as a progressive brain disease. Schizophr Bull. 2012. Abstract
View all comments by Robert B. ZipurskyComment by: S. Charles Schulz (Disclosure)
Submitted 6 September 2013
Posted 6 September 2013
This study by Palaniyappan et al. is an excellent step in using neuroscience measures—in this case MRI imaging—to address the important issue of treatment outcomes. As noted by the authors, this is a new way to address outcome in the early phase of psychotic disorders and may be related to the issue of poorer outcome with increased duration of untreated psychosis (DUP). In other words, the literature does show that approximately 25 percent of young people with schizophrenia are not responsive to first-line antipsychotic medication (Agid et al., 2011). Therefore, ways to provide the best treatments sooner can be very helpful in improving outcome.
The authors use an excellent test and examine more than just a single brain area. This newer strategy can lead to much better assessments than earlier work with a single measure. Further, by using MRI measures in the first episode of psychotic illness, they are adding to the initial evaluation of psychotic illness—making sure that there is not some other cause of psychosis.
In addition, it is very useful that they examined a range of psychotic disorders and note that the assessment is useful across these illnesses; therefore, the findings are not limited to schizophrenia alone. Some clinicians will be aware that this may lead to a broader assessment of treatment initiation and moving along on algorithms for a broader range of patients.
In summary, this is a very good contribution and can lead to new care pathways in the early stages of psychosis.
Agid O, Arenovich T, Sajeev G, Zipursky RB, Kapur S, Foussias G, Remington G. An algorithm-based approach to first-episode schizophrenia: response rates over 3 prospective antipsychotic trials with a retrospective data analysis. J Clin Psychiatry. 2011 Nov;72(11):1439-44. Abstract
View all comments by S. Charles Schulz