Abstract
It has long been known that intracranial tumours are associated with a high incidence of psychiatric symptoms, including depression, hallucinations, personality changes, emotional disturbances, and intellectual failure [1–3]. Parry [4] reported that 1/200 patients admitted to a psychiatry unit had a brain tumour. The largest series was reported by Keschner et al. [5] who found mental symptoms in 94% of cases with tumours of the temporal lobes, in 90% of cases with neoplasm of the frontal lobes, and in 47% of cases with infratentorial tumours. In a hospitalized psychiatric patient population at the University Clinic in Basle, Kocher et al. [6] found primary brain tumours in 1/1000 patients, a rate approximately 20 times that of the general population. These studies do not discuss the differential prevalence of psychiatric symptoms separately for benign and malignant tumours.
Despite the fact that psychiatric morbidity is relatively common in patients with certain brain tumours, it is infrequent for a psychiatrist to discover cerebral tumours in their patients since brain scans are not routinely ordered. In a prospective study, Hollister and Butros [7] ordered 337 CT or MRI scans to examine the frequency of undiagnosed conditions. They found that 28% of the patients had CT or MRI scans that were abnormal. Only two patients had brain tumours, which changed their management from psychiatric to neurosurgical.
Meningiomas are the most common benign brain tumours in adults and as such are encountered fairly often in neurosurgical practice [8]. The prevalence of meningiomas varies from 2.3 cases per 100 000 during life for the general population, to 5.5 per 100 000 if autopsy data are included. The incidence rates of the meningiomas were 1.5 and 3.1 per 100 000 for men and women, respectively, and the rates rose with increasing age.
In order to examine the psychiatric presentation associated with meningiomas, we undertook a retrospective hospital case record analysis of a consecutive series of all inpatient admissions to the departments of neurology and neurosurgery over a period of 5 years (January 1996– March 2001) at the Canberra Hospital.
Method
The Canberra Hospital is a university general and speciality hospital with a bed capacity of 500, serving a population of approximately 340 000 in the region of the Australian Capital Territory and neighbouring New South Wales, Australia. This is an exclusive neurosurgical centre for the catchment area unless patients are referred to Sydney or other major centres.
The diagnoses of brain tumour were based on the record of final diagnoses in the case records confirmed by either CT or MRI scans. Case records of patients with clearly documented history of psychiatric symptoms of several weeks to several months duration were identified if such symptoms had antedated a diagnosis of brain tumour. Using a specially designed proforma, two psychiatrists rated the symptoms in a joint session. We also collected data on age, gender and CT/MRI findings. Consensus was reached on all cases in regard to the psychiatric phenomenology. The symptoms were divided according to their presentation into purely neurological, purely psychiatric and combination of both neurological and psychiatric. A patient was considered to suffer from anxiety or depression if the case notes recorded these terms in the context of a GP diagnosis or the need to treat such symptoms. Similarly personality change was defined as any observation by the family members of any abrupt change (e.g. apathy, aggressiveness or irritability) in established behaviour patterns. Admittedly DSM-IV or ICD-10 diagnosis was not possible in a retrospective study. The combination of neurological and psychiatric symptoms was considered to be a neurological presentation for the purpose of this analysis. In the event of ambiguity the symptoms were considered to have a neurological presentation and not psychiatric.
Results
A total of 79 patients were identified to have a primary diagnosis of benign brain tumours. There were 56 female patients and 23 male patients. Seventy-two of these patients had meningiomas. Three men and four women had non-meningioma tumours.
Of the 72 with meningioma, 15 patients (21%), eight men and seven women presented with psychiatric symptoms in the absence of neurological symptoms. The remaining 57 patients with meningioma, and all of the seven with non-meningioma had presented with clear neurological symptoms.
Five of the eight men and all the women who presented with psychiatric symptoms suffered with affective symptoms, taking the form of symptoms of depressive illness and/or anxiety, including one man who developed treatment-refractory depression and two patients with psychotic depression. Three men presented with personality related changes.
The mean age for our patients with psychiatric symptoms and meningiomas was 60.5 years (SD ± 16.2). This is the same as for the whole group of meningioma patients.
An analysis was conducted in regard to the location of lesions (hemispheric as well as anatomical sites) and the results are shown in Table 1.
Location of lesions
Discussion
A major methodological difficulty in this study as with many retrospective studies is that of ascertainment of psychiatric symptoms in patients' case notes. Thus it was not possible for us to allocate a DSM-III, DSM-IV or ICD-10 diagnosis.
Overall 21% of patients with meningiomas in our series were found to have psychiatric symptoms; 35% of the men who had meningiomas had presented with psychiatric symptoms as compared to 15% of women. Psychiatric comorbidity with meningiomas has been previously reported in sporadic case reports as well as in two recent studies. Lampl et al. [9] found similar rates as us of psychiatric symptoms among meningioma patients. Using DSM-III-R criteria they demonstrated psychiatric comorbidity in 16 of the 50 meningioma patients. Among these, nine had major depression, four atypical depression and three unspecified psychosis. Pringle and Whittle [10] reported higher levels of anxiety and depression in patients with meningiomas when measured on the Hospital Depression Anxiety Rating Scale compared to other intracranial neoplasms. These patients were assessed after a mean of 6.7 days of receiving a radiological diagnosis of intracranial neoplasm thus confounded by a psychological reaction to the diagnoses received. In our study, 80% of the patients who had meningiomas and had psychiatric symptoms had presented with affective symptoms taking the form of depressive illness or anxiety. Nonetheless the observation that affective symptoms, depression and anxiety are common to the presentation of meningiomas is in line with the reports of Lampl et al. [9] and Pringle and Whittle [10].
Lampl et al. [9] also reported the psychiatric comorbidity in their series and was limited to patients with right hemispheric frontal lobe meningiomas. In our study psychiatric symptoms occurred in seven of the 15 patients with meningiomas on the left side of the brain and only three patients with meningiomas on the right side. Five patients had tumours in the midline or with a bilateral location. Of two patients who had personality related changes, one had a left frontal lobe meningioma and the other a left temporal lobe meningioma. Pringle and Whittle's cohort also appear to have a bilateral, right as well as left side, distribution of the meningiomas implicated in causing higher scores on the Hospital Depression Anxiety Rating Scale. In our data none of the patients with psychiatric symptoms had a meningioma of the base of the skull. Lampl et al. have observed that none of their patients with psychiatric comorbidity had meningioma at the base of the skull compared to 44% of the brain convexity meningiomas presenting with psychiatric comorbidity. Thus it would appear the laterality of brain tumour site and the psychiatric comorbidity has no definite positive correlation.
We found that for the general population and the period covered by this study, the number of meningioma cases recorded at our hospital give an overall incidence of 4.2 cases per 100 000 per year. This incidence is similar to the incidence reported in the Manitoba study [11]. Dumas Duport [12] observed ‘meningiomas tend to produce chronic pictures of mental disorder and therefore tend to be missed’. The higher representation of women among all meningioma cases is consistent with the higher incidence of meningiomas in the female population in general. Although our patient population demonstrated a higher incidence of meningiomas in women than men, the number of such women with psychiatric symptoms was similar to the number of men with psychiatric symptoms. Whether this reflects a lesser chance for women with meningiomas to develop psychiatric morbidity compared to men, will need to be resolved by future, larger surveys.
According to Ron [13], there is no raised incidence of brain tumours among the psychiatrically ill, and therefore, the need for excessive caution and the routine expenses of clinical investigations should be spared. Weitzner [3] claims primary brain tumours are increasing in incidence with considerable psychosocial impact due to a direct effect of the tumour on mood, personality, cognition, perceptions, and indeed upon quality of life, as well as life expectancy. The difference between live population incidence of meningiomas and the autopsy incidence of meningiomas indicates many meningiomas remain undiagnosed and untreated. Our data support the conclusion that ‘affective symptoms’ and or ‘personality change’ occur in a considerable proportion of patients with meningiomas. The mean age (60.5) for the population of patients in this study with meningiomas who presented with psychiatric symptoms did not differ from that in the overall population of meningioma patients in our series. Lampl et al. [9] data refers to similar age finding. Holister and Boutros reach the conclusion that a first psychotic episode or personality change after the age of 45 years should be a sound indication for CT or MRI brain imaging in psychiatric practice. Of the 72 patients with meningiomas, 15 patients with psychiatric symptoms in our study were over the age of 45 and this lends support to this suggestion.
We conclude that psychiatric presentations are common with meningiomas occurring in the fourth decades of life and these patients may present to psychiatrists for treatment of their psychiatric problems. Such patients should be screened by CT scan or MRI to rule out benign tumours of the brain, particularly meningiomas.
Footnotes
Acknowledgements
We thank Jaya Amsterdam and David Brunswick from the Depression Research Unit of the University of Pennsylvania in Philadelphia, for prompting the study and in the preparation of the manuscript.