Heterogeneity of response to antipsychotics from multiple disorders in the schizophrenia spectrum.

BACKGROUND Antipsychotic response after the initiation of neuroleptic treatment shows wide variation in schizophrenic patient populations. In this overview, the authors suggest that the variance in antipsychotic drug response within schizophrenia can be reduced by resolving the schizophrenias into several discrete "endophenotypes," each with different etiologic underpinnings. METHOD Studies relating differences in the relative speed or completeness of antipsychotic response to differences in distribution of 2 biological markers with possible etiologic significance are reviewed. Such studies had assessed recently hospitalized, neuroleptic-free patients undergoing exacerbation of nonaffective psychotic disorders. Prior to initiation of neuroleptic, the cohort of patients had been assessed for the quantity of the dopamine metabolite homovanillic acid in plasma (pHVA) and had undergone the first of 2 magnetic resonance imaging (MRI) studies for analyses of ventricle volumes. A second MRI was subsequently performed during a period of (partial) remission to determine within-patient stability of ventricular volumes. These selected studies assessed the distribution of pHVA and distribution of rates of ventricular change, with non-normal distributions resolved by K-means clustering. The speed and completeness of neuroleptic-induced antipsychotic response were related to 3 clusters of patients delineated by modal distributions of pHVA and of apparent rates of ventricular change. RESULTS At least 3 unique "endophenotypes" of the "group of the schizophrenias" can be defined with respect to speed and completeness of antipsychotic response. Each endophenotype appears to show at least one unique biological feature that differentiates it from a normal comparison group. A rapidly responsive psychosis was associated with excessive production of dopamine, as identifiable by elevation of pHVA and a "good-prognosis" course. A delayed-response psychosis had low-to-normal pHVA, clinically demonstrated persistent negative symptoms, and was associated with an excessive rate of change in ventricle volume between exacerbations of psychosis and (partial) remissions. Finally, a nonresponsive psychosis could be characterized as having both low-to-normal pHVA and rate of change of ventricle volumes similar to that of controls. Additional studies revealed that each of the endophenotypes had high rates of the psychoses in family members. The good-prognosis course of the rapidly responsive group of studied patients was also found in their family members who had psychotic disorders. Similarly, the prominent negative symptoms of the delayed-response probands were reflected as a prominent trait in their family members also afflicted with psychosis. The endophenotypes tended to "breed true" in terms of prognosis and negative symptoms. CONCLUSION Major differences in antipsychotic response patterns appear to be associated with patient and family characteristics that may be related to differences in the etiology and consequent pathophysiology of illness.