Secondary Schizophrenia (105 page)

about 3% as compiled from autopsy series from psy-typically involve nonauditory hallucinations,
chiatric hospitals
[5].
In fact, the existing literature
neurological signs, and prominent cognitive
is inconsistent as to whether the prevalence of brain
impairment.

tumors is greater among psychiatric or general hospi-r
Tumors presenting with psychotic symptoms
tal patients
[6]
. Nonetheless, the presence of a brain
are most likely to be in the frontal or

tumor in this group of patients with an established
temporal lobes or in the pituitary.

psychotic disorder appears to be important to recog-r
nize given its potential to confound the presentation,
The most common association is with

management, and outcome of the primary psychotic
low-grade, slow-growing tumors such as

disorder.

meningiomas and low-grade gliomas.

r
Treatment directed at the tumors, such as the
use of steroids, may sometimes provoke

Definitions and diagnostic criteria

psychotic symptoms.

r
It is important for psychiatrists to be
Tumor classification

“brain-tumor conscious.”

Brain tumors may be classified according to whether
they are primary or metastatic or according to

Brain tumors most commonly present with features of
mary tumors comprise about 80% of these, whereas
raised intracranial pressure, focal neurological signs,
metastatic tumors contribute to the remaining 20%

or seizures
[1]
. However, the earliest manifestation of
[7]
. Seventy percent of all intracranial tumors are
some intracranial tumors may consist of psychiatric
located supratentorially, with most situated in the pos-symptoms or behavioral disturbance alone
[2].
More
terior fossa and temporal and frontal lobes
[7].

specifically, psychosis may be the sole presenting fea-The commonest primary tumors are gliomas with
ture, increasing the likelihood of such patients com-meningiomas; pituitary adenomas are the next most
ing to the attention of a psychiatrist. Although psy-frequent
[7].
Low-grade gliomas (grade I–II) are
chosis secondary to brain tumor is relatively rare
[3],

slow growing and generally benign whereas glioblas-this subgroup is at risk of misdiagnosis and delayed
tomas (grade IV) are rapidly growing and aggres-treatment, potentially resulting in increased morbidity
sive
[1].
Metastases occur most frequently secondary
263

and mortality. A 30-year retrospective study of psychi-to primary lesions in the lung and breast
[7].
Brain
Organic Syndromes of Schizophrenia – Section 3

metastases, especially if multiple, are associated with
characterization of psychotic symptoms was also varia poor prognosis with average survival rates between 1

able, with older studies often grouping together psy-to 2 years
[7].

chotic, affective, and acutely confused states.

With most studies adopting a retrospective study
Diagnostic criteria

design, it has been difficult to establish causality between brain tumors and psychotic symptoms.

Psychosis secondary to brain tumor is included within
Autopsy studies have typically not differentiated
the broader classification of “psychosis due to a gen-between brain tumors producing psychiatric symp-eral medical condition” according to the Diagnostic
toms and tumors developing in patients with pre-and Statistical Manual of Mental Disorders (DSMMD)
existing psychiatric illness, increasing the likelihood
[8].
This stipulates that psychotic symptoms must be
that a coincidental finding may be misattributed as
identified as a direct physiologic consequence of the
causation.

medical condition. The challenge emerges in demon-Despite the limitations of the existing epidemiolog-strating this causality in the context of confounding
ical data, some general conclusions may be drawn from
factors such as the presence of a pre-existing psy-the heterogeneous case material.

chiatric or neurodegenerative condition and specific
The prevalence of cerebral tumor in the general
treatment for the brain tumor (e.g., steroids, radio-population is 0.16%
[12],
whereas the prevalence of
therapy). This diagnosis necessitates the exclusion of
intracranial tumors at autopsy is 1 to 1.5%
[13].
This
a primary psychotic disorder, substance-induced psy-significant difference is presumably because many
chosis, dementia, and delirium as better diagnoses
tumors remain neurologically silent throughout life
to account for the psychotic symptoms. The clinical
and may only be discovered at autopsy. Despite some
features of the psychoses concerned are similar to
variation in figures, the prevalence of intracranial
the research diagnostic criteria for schizophrenia pro-tumors in psychiatric patients is consistently reported
posed by Spitzer and colleagues
[9].

to be higher. A review of nine mental hospital autopsy
More broadly, psychosis secondary to a brain
studies between 1909 and 1949 quoted figures of 1.7 to
tumor has been accepted as an “organic psychosis,” as
11.2%
[14],
whereas other studies reported figures of
opposed to a “functional psychosis”
[10].
However, the
up to 13.5%
[15, 16].

term “organic” is problematic as it implies the pres-In comparison, it is extremely common for patients
ence of “identifiable structural brain disease or toxic-with cerebral tumors to display psychiatric symp-metabolic brain dysfunction,” which has been shown
toms. The largest series of tumor patients studied by
to be the case in schizophrenia also. Hence, the most
Keschner and colleagues
[17]
reported mental symp-recent alternative classification proposes “primary and
toms in 78% of 530 cases. However, very few studies
secondary schizophrenias”
[11].

solely reported the incidence of psychotic symptoms.

In their study of the literature, Lisanby and col-

Epidemiology

leagues
[6]
found that tumors presenting with psy-The prevalence of intracranial tumors associated with
chiatric symptoms tend to be more common among
psychiatric symptoms is difficult to estimate, with
the elderly. The peak age of diagnosis was the seventh
the existing literature yielding inconsistent findings.

decade, with 27.6% of cases falling into this age range.

There are a number of methodological problems with
An excess of meningiomas in particular appear
many of the studies, including the high variabil-to be found in psychiatric patients compared to the
ity and poor sensitivity of screening methods. The
general population in autopsy studies
[18, 19].
Pat-nature of the populations studied has been highly vari-ton and Sheppard
[18]
found meningiomas consti-able (e.g., autopsy vs. living patients, psychiatric vs.

tuted 33.2% of tumors in psychiatric hospitals com-general hospital patients). Furthermore, the concep-pared with 13.7% in general hospitals. Rapidly pro-tualization of both psychiatric and neurological ill-gressive gliomas are more likely to present with focal
nesses has changed over the period these studies have
neurological signs whereas slowly progressive menin-been done. Older studies were more likely to com-giomas may be more likely to present with psychi-bine all brain lesions associated with mental symp-atric symptoms in advance of any neurological signs.

toms, including those of nonneoplastic etiology such
In a 5-year retrospective data analysis of benign brain
264

as subdural hematomas and infectious lesions. The
tumors, Gupta and colleagues
[20]
found 21% of
Chapter 20 – Brain tumors

meningioma cases presented with psychiatric symptoms more often than other tumors
[3]
. More recently,
toms in the absence of neurological symptoms.

Davison and Bagley
[26]
specifically reviewed the literature on patients with psychosis secondary to brain
Clinical manifestations

tumors and found no predominance of one tumor type
over another. Lishman
[2]
has suggested that menin-Association with tumor location and
giomas and low-grade gliomas are more likely to produce behavioral and psychiatric symptoms because of
laterality

their slow-growing nature, which may delay earlier
The significance of tumor location in relation to mental
neurological signs. Mostly, however, it appears that
symptoms has been much debated. Although tumors
tumor histology is less significant than other factors
in many areas have been implicated in the causation of
in determining the presence and nature of psychiatric
psychosis, no areas, when lesioned, have reliably pro-symptoms.

duced psychotic symptoms
[21].

Psychosis has been reported to be associated with
Other tumor features influencing

intracranial tumors in a wide variety of locations,
including the temporal lobes
[22, 23, 24, 25, 26, 27,

symptom formation

28, 29, 30,
31],
frontal lobes
[31, 32],
suprasellar region
Price and colleagues
[45]
list the extent of tumor
[26, 28,
31, 33],
diencephalons
[25, 26,
34, 35, 36, 37,

involvement, the rapidity of its growth, and the
38, 39, 40, 41],
corpus callosum
[22],
pituitary
[3, 26,

propensity to cause raised intracranial pressure as
28,
31, 42],
and posterior fossa
[22,
43].

the most significant factors influencing symptom for-Frontal- and temporal-lobe tumors have consis-mation. Other reports, however, suggest increased
tently shown a higher incidence of psychiatric distur-intracranial pressure as less likely to be a consistent
bance compared with parietal, occipital, and infraten-major factor
[2, 28].

torial tumors
[2, 17, 25, 26,
44, 45].
More specifi-In addition, the patient’s psychiatric history, level
cally, psychosis is most common with tumors affect-of functioning, premorbid personality, and coping
ing temporal lobe and limbic structures
[21,
46].

mechanisms may contribute to the nature of an indi-Mulder and Daly
[23]
reported that 20% of tumors
vidual’s response to a cerebral tumor and the develop-affecting the temporal lobes were associated with
ment of psychotic symptoms
[45].
These highly indi-schizophrenia-like psychosis. In a series by Malamud
vidual host factors may modify the psychiatric effects
[25],
schizophrenia-like psychosis was found to be
of an intracranial tumor wherever it is situated.

associated with tumors in the hippocampus, amyg-dalae, and cingulate gyrus. Davison and Bagley
[26]

Nature of psychotic symptoms

found that schizophrenia was significantly associated
with pituitary tumors and that temporal lobe and
Psychotic symptoms secondary to brain tumor may be
suprasellar tumors were specifically associated with
indistinguishable from symptoms due to a functional
hallucinations.

psychotic disorder. In particular, patients with tempo-In terms of laterality, case reports, although not
ral lobe tumors have been noted to have a high fre-larger series, show a preponderance of left limbic
quency of schizophrenia-like psychotic features
[23,

tumors
[27, 29].
Some of these reports have suggested
24, 25].

that psychosis may remit when left-temporal tumors
However, some general differences have been
have been excised.

noted when comparing the psychotic symptoms secondary to brain tumor to those of schizophrenia. Cummings
[47]
emphasized that delusions secondary to
Association with tumor type

brain tumor tend to be simpler in nature than the
There is no consistent evidence to suggest that the
systematized delusions of schizophrenia
[47].
These
histological type of the tumor is linked to the pres-delusions are most commonly persecutory or para-ence or nature of psychotic symptoms
[6, 17].
Consis-noid in nature
[6, 28].
Visual hallucinations are the
tent with the above-quoted literature related to broader
commonest type of hallucination secondary to brain
psychiatric symptomatology, early reports suggested
tumor, followed by auditory hallucinations
[28].
This
265