By William T. Carpenter, MD
[Winter 2006; Vol. 32, No. 2; Pg 11, 13]
Schizophrenia:
From
Syndrome to Pathologic Processes to Disease
This
is a nice opportunity for me to provide Maryland Psychiatric Society members
with a brief overview of an aspect of my research on schizophrenia in which my
colleagues and I have been involved since 1968.
This work is based on the assumption that schizophrenia is a clinical
syndrome with more than one disease process and heterogeneity in
etiopathophysiology among cases.
The
starting point involved data from the International Pilot Study of Schizophrenia
(IPSS) collected in nine nations beginning in 1968.
The concept of nuclear schizophrenia defined by Schneiderian first rank
symptoms and Langfeldt’s distinction between true and pseudoschizophrenia was
dominant in Europe, and was highly influential in the DSM III process.
We first showed that patients with schizophrenia as defined by Schneider,
Langfeldt, or our 12 point flexible system were not distinguished from
schizophrenia subjects not meeting nuclear schizophrenia criteria on prognostic
or outcome variables; thereby falsifying the nuclear schizophrenia hypothesis.
DSM III, nonetheless, created a definition of schizophrenia based mainly
on reality distortion pathology.
This definition moved substantially away from Kraepelin’s emphasis on
avolitional and dissociative pathologies and Bleuler’s assertion that
dissociative thought was the core pathology.
While
conducting the Washington component of the IPSS, we created scales to measure
multiple aspects of pre-psychotic development and post-psychotic course of
illness.
This had the effect of changing the onset and course focus from psychosis
and relapse to a multifactorial construct.
These data enabled us to discover semi-independent domains of pathology
within the schizophrenia syndrome.
We proposed a tripartite model comprising positive psychotic symptoms
combining disorganization of thought and reality distortion (i.e., delusions and
hallucinations); pathology comprising diminished normative functions termed
negative symptoms; and pathology observed in interpersonal function such as
impaired rapport.
We
found these domains to have little within subject relation at any one point in
time, and virtually nil relationship over time.
On the other hand, pathology within each domain was closely related to
past pathology in that domain, and predictive of future pathology within the
same domain.
In 1974 we proposed that each domain be investigated as a separate
pathologic process with potentially different etiology and clearly different
treatment implications.
This proposal challenged the dominate single disease paradigm where the
various manifestations were assumed to arise from the same disease process.
Rather, the tripartite model presumed separable neuropathologic
substrates for various symptom domains.
The dominant model, with psychosis as proxy for the disease class, has
led to a narrow discovery process regarding etiology, pathophysiology, and
treatment.
The alternative model, requiring focus on domains of pathology instead of
diagnostic class, has increasing influence as seen in the current emphasis on
negative symptoms and impairments in cognition.
In
the next step, we tested the hypothesis that negative symptoms represent a
separate pathologic domain within schizophrenia.
Since many patients are diagnosed with schizophrenia based only on
positive symptoms, we considered a comparison of schizophrenia with negative
symptoms to schizophrenia without negative symptoms to be the ideal design.
Many sources of artifact, such as being on antipsychotic drugs, would be
balanced between groups in these comparisons.
From
a series of studies at the MPRC, we have reported many substantial differences.
These include pre-psychotic development; clinical features such as drug
abuse, depression and social content of delusions; and a number of neuroimaging
and electrophysiologic differences.
Consistent with the domains construct, variables associated with
psychosis were similar between groups.
This included the response of psychosis to antipsychotic drugs.
However,
negative symptoms, if primary to schizophrenia, have not proven responsive to
antipsychotic drugs including clozapine. PET and MRI studies revealed that the
neural substrate for negative symptoms is in the inferior parietal lobe and
dorsolateral prefrontal cortex.
The pattern of structural change suggests smaller volume of hippocampal
and prefrontal white matter tissues in cases without primary negative symptoms;
and enlargement of other structures [i.e., caudate and lobules VIII-X of the
cerebellum] in the subgroup with primary negative symptoms—a group we refer to
as deficit schizophrenia.
This was particularly impressive in a study where normal volunteers and
non-deficit patients had similar volumes in this cerebellum structure, and a
positive correlation between size and cognitive function.
In contrast, this structure was larger in deficit schizophrenia with a
negative correlation between size and cognitive function.
The
clinical studies adumbrated above confirm that the negative symptom domain is
importantly distinct from psychosis.
Using traditional criteria for identifying a disease entity [i.e.,
pattern of onset, clinical manifestations, biologic correlates, treatment
response and course], we hypothesized that negative symptoms were not merely a
separate domain, but marked the presence of a different disease.
This hypothesis would be supported if etiologic factors differentiated
deficit and non-deficit schizophrenia.
Two risk factors have been investigated with leadership from Brian
Kirkpatrick.
First, the pattern of morbid risk in pedigree studies appears different.
Compared to non-deficit probands, relatives of deficit schizophrenia
probands are more likely to have schizophrenia, the deficit form of
schizophrenia, and social isolation in the absence of psychosis.
If the proband has non-deficit schizophrenia, there is an increased risk
[compared to deficit probands] of affective psychoses and mental disorders
generally. Second, deficit schizophrenia patients are substantially more likely
to be born during the summer months compared with non-deficit cases and
population norms.
This has been repeatedly replicated and confirmed in meta-analysis.
This brief presentation does not cover our
therapeutic studies or work in ethics except to note that there is now general
agreement that we do not have efficacious treatment for primary negative
symptoms or impaired cognition. But
the work in psychopathology leads to several conclusions.
Boldly stated, four outcomes are:
Falsifying the nuclear schizophrenia construct
and damaging the single disease paradigm;
Introducing the tripartite model and showing the within domain consistency over
time and the between domain independence over time;
Shifting the schizophrenia paradigm to one based on domains of pathology; and
Meeting, for the first time, the 100 year challenge [e.g., Blueler’s “the
group of schizophrenia”] of identifying a disease entity within the
schizophrenia syndrome.
Immediate payoff includes emphasis on novel
treatment discovery for each domain and the search for genotypes associated with
domains rather than schizophrenia as a homogeneous class.
NIMH has now made this agenda a top priority.
We recently became responsible for editing and publishing the
Schizophrenia Bulletin, and these NIMH initiatives are described in the January
and October, 2005 and April, 2006 issues.
[Commercial note: subscriptions are
inexpensive and available online at http://schizophreniabulletin.oxfordjournals.org/]
For me personally, it has been a
wonderful experience to work with such terrific colleagues and friends, John
Strauss and John Bartko in the early work, and many scientists at the MPRC in
the work since 1977. Bob Buchanan and Brian Kirkpatrick have been involved in
all aspects of this work, giving leadership to much that we have accomplished.