For the most part, drugs in psychiatry have been targeted at psychiatric disorders that are categorized in diagnostic manuals such as Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition
. This has been accepted by regulatory authorities such as the FDA and has guided the development of drugs for psychiatric disorders as well as agents in other areas of medicine. A 2003 article by Hyman and Fenton13
questioned whether this was the best strategy for drug development. Because researchers attempt to connect human phenotypes as defined by diagnostic category to genetics, the limitations of this method have become more apparent. Psychiatric disorders tend to have complex phenotypes with patients demonstrating abnormalities in multiple areas of brain functioning. This is particularly true in schizophrenia, a disorder that includes psychopathology in different dimensions. Although researchers continue to debate about the number of dimensions, there is agreement that individuals with schizophrenia are vulnerable to having impairments in their perceptions of reality (or psychotic symptoms), in their cognitive abilities, and in their basic drives (or negative symptoms).14
The proposal of Hyman and Fenton13
suggests a strategy for advancing drug development by dissecting Diagnostic and Statistical Manual of Mental Disorders
into component dimensions of psychopathology that may be more proximate to pathophysiological processes. These more circumscribed areas of psychopathology may cut across diagnostic boundaries. For example, the pathophysiology of cognitive disturbances in schizophrenia and Alzheimer disease may have similarities and so may the apathy in schizophrenia and major depression. Therefore, pharmacological approaches to these dimensions may cross diagnostic boundaries. This novel approach to addressing psychopathology is actually reflected in clinical practice. As mentioned above, antipsychotic medications are not antischizophrenia medications. Rather, they are effective agents for any type of psychosis. Antidepressant medications are not just effective for major depressions. They are effective for depressive symptoms in a number of psychiatric disorders.
This more pragmatic approach to drug development suggests a number of promising targets. Hyman and Fenton13
viewed the cognitive impairments in schizophrenia as particularly promising. These impairments affect nearly every individual with schizophrenia to some degree.15,16
Moreover, the severity of these impairments is related to the ability of individuals with schizophrenia to function in the community.17
They raise the very compelling question as to whether a drug that improved cognition might, in turn, improve functioning. It is important to note that this proposal by Hyman and Fenton13
has helped to reorient treatment development. As will be noted below, Fenton was the individual, who demonstrated extraordinary leadership within NIMH by translating these ideas into effective Institute initiatives.
An interesting byproduct of the current process of drug discovery relates to the development of clinical measures. The drugs that treat schizophrenia are reasonably effective for attenuating or even eliminating psychotic symptoms. As a result, clinical investigators have learned to measure hallucinations, delusions, and other psychotic symptoms with good accuracy. Dimensions of psychopathology that are less affected by antipsychotics—particularly cognitive impairments and negative symptoms—can be measured using a number of instruments. However, it is unclear how these instruments are when they are used as endpoints in clinical trials. Instruments that are reliable and valid for measuring psychosis such as the Positive and Negative Symptom Scale18
and the Brief Psychiatric Rating Scale19
are widely accepted within the academic community, in industry, and by FDA. Until recently, there was little agreement on the best instruments for measuring cognition and negative symptoms.
The NIMH-MATRICS initiative was a direct result of the strategy proposed by Hyman and Fenton.13
The goal of this program (awarded as a competitive contract to UCLA with Stephen Marder and Michael Green as coprincipal investigators) was to facilitate the development of pharmacological treatments for the cognitive impairments that are common in schizophrenia by addressing key obstacles. These obstacles included the lack of agreement regarding how to measure cognition in a clinical trial; concerns about the willingness of regulators such as the FDA to accept cognition in schizophrenia as a drug indication; a lack of agreement regarding the best molecular targets for drug development; and concerns about the clinical trial design that would be appropriate for demonstrating the effectiveness of a drug. During the 2-year period from 2003 to 2004, 6 MATRICS conferences addressed all these issues. (Each meeting is described at http://www.matrics.ucla.edu
The MATRICS Consensus Cognitive Battery (MCCB) may be the most tangible result of this initiative. A committee of experts cochaired by Michael Green and Keith Nuechterlein first achieved agreement about the domains of cognition that were impaired in schizophrenia and which should be the focus of measurement.20
This was done through a process that included both expert opinion and factor analysis. The committee also achieved agreement regarding the selection criteria for tests within each domain. Using a process developed by RAND to measure the appropriateness of medical treatments,21
the group convened a panel to compare each candidate test with the selection criteria. This led to the development of a beta version of the battery, which included 20 of the most promising tests. This version was evaluated in a 5-site study, where the battery was first administered to patients with schizophrenia on 2 test occasions separated by a month. The battery was then administered to a large community sample to collect normative data.
At this time, the MCCB is the only battery that has been accepted by FDA as an instrument developed through broad consensus building. For the convenience of researchers, the battery has been assembled into a single kit, which can be purchased through major test publishers. The battery is currently being used in a number of large studies of putative cognition enhancing drugs.
A similar effort was used by MATRICS to achieve agreement on the most promising molecular targets for drug development. A Neuropharmacology Committee cochaired by Carol Tamminga and Mark Geyer surveyed experts in the field regarding promising targets. The survey results (available at http://www.matrics.ucla.edu
) were used to develop the agenda for a meeting that was held at the NIH Clinical Center on June 23–24, 2003. At the meeting, experts presented evidence supporting each targets. The proceedings of the meeting were assembled into a special issue of Psychopharmacology in 2004.22
A recent review also provides an overview of the targets recommended by MATRICS.23
One of the important accomplishments of MATRICS was its ability to organize important interactions with FDA. Individuals from FDA participated in a number of live MATRICS meetings and conference calls. In addition, MATRICS organized an NIMH-FDA workshop on clinical trials methodology. At the first MATRICS meeting, a representative from FDA, Dr Thomas Laughren, agreed that impaired cognition is a core feature of schizophrenia and that if a drug demonstrated that it could improve cognition it could be approved. In April 2004, NIMH-FDA representatives from both industry and academia as well as government reached agreement on clinical trial designs that could support an efficacy claim for either a comedication that could be added to an antipsychotic to improve cognition or an agent which was effective for treating psychosis and improving cognition.21