Abstract
Major psychopathology has been well described among patients with human immunodeficiency virus (HIV) infection and acquired immune deficiency syndrome (AIDS) [1]. Mania, in particular, is reported to be significantly increased in prevalence in AIDS patients when compared with the general population [2–A].
Mania typically occurs as part of bipolar affective disorder (manic depressive illness), but it may also occur secondary to a variety of medical and pharmacological antecedents.
Recognising the possible subtypes of mania has implications for aetiology, therapy, and prognosis of the illness [5].
Krauthammer and Klerman suggested mania should be considered a heterogeneous syndrome, most usefully subdivided to primary and secondary mania [6]. Their suggested criteria to identify secondary mania are: close temporal proximity of an organic insult to the subsequent mania; negative premorbid history; predominantly negative family history; and late age of onset.
Subsequently, a number of reviews and case reports have expanded the range of identified medical conditions and drugs associated with mania. Among the more common causes reported are neurological conditions, especially brain injury [7, [8], [9], [10], [11], [12], [13]], cerebrovascular disease [14, [15], [16]], and multiple sclerosis [17, [18]]. General medical conditions reported include hyperthyroidism, polycythemia and infections [6, [19]]. Drugs associated with secondary mania include steroids, acyclovir, isoniazid, antidepressants and amphetamines [6, [19], [20]].
Mania and HIV infection
The first report of mania associated with HIV infection was published in 1984. This described a 31-year-old homosexual male who had a 6-month history of Kaposi's sarcoma prior to presenting with Pneumocystis carinii pneumonia (PCP). Within 2 weeks of treatment for PCP, he developed symptoms of an acute brain syndrome and experienced a seizure. Following the seizure he developed symptoms of mania with psychotic features and confusion. Treatment with haloperidol was successful [21].
Since this report, there have been numerous case reports and five case series describing mania in patients with HIV infection [2, [3], [4], [22], [23]].
Case studies documenting episodes of mania in patients with HIV infection fall into three groups: those in which HIV infection appears to be the only organic factor potentially contributing to the development of mania [24, [25], [26], [27]]; those in which mania appears secondary to antiretroviral medication (e.g. zidovudine [28, [29], [30]] and didanosine [31]); and those in which mania developed secondary to HIV-related conditions (e.g. cryptococcal meningitis [32]) or other medications (e.g. antidepressants [33, [34]]). Among cases of HIV-related mania some authors [3, [23], [24], [25], [26], [27]] noted the occurrence of concurrent or subsequent cognitive impairment, while others [23, [27], [35]], report mania in the absence of any cognitive impairment.
Of the five series of cases describing HIV-related mania, three identified cases for study by retrospective chart review [2, [3], [4]] and two identified cases from a cohort referred for neuropsychological or psychiatric evaluation [22, [23]]. Only two series [2, [23]] delineated a group that could be defined as secondary mania according to Krauthammer and Klerman (presence or absence of family and personal history of mood disorder). The first of these studies [2] demonstrated that patients with secondary mania were more likely to have a low CD4 count (> 100 cells), to have developed AIDS and to have a comorbid diagnosis of dementia. The authors note that the 17-month prevalence of mania in their population of HIV-positive patients was 1.4%, and the rate in the patients with AIDS was 8%. This prevalence is 10 times the 6-month prevalence expected in the general population.
The second study [23] extended these findings, with further evidence suggesting a link between HIV-related mania and dementia, and also provided evidence that the clinical features of ‘HIV mania’ differed to those of primary mania (patients with HIV mania had more symptoms, and were more likely to display irritability, and less likely to display increased talkativeness).
With the exception of the studies by Lyketsos and colleagues [2, [23]], no attempts have been made to reliably identify a cohort of patients with secondary mania and to examine the features of this disorder. This pilot study was undertaken to identify a cohort of patients with mania secondary to HIV infection, to describe the clinical and neuroradiological features of HIV-related mania, and to describe the treatment outcome of the patients.
Method
This study was undertaken at Fairfield Hospital, which was a 138-bed infectious diseases hospital in Melbourne. Forty beds were delineated for the treatment of HIV/AIDS patients. During the study period, Fairfield Hospital was the major service for the treatment of HIV/AIDS in Victoria, caring for 70% of all patients diagnosed with HIV/AIDS at that time.
All patients referred to the HIV consultationliaison (C–L) psychiatry service over the 29-month period from January 1993 to June 1995, were screened for the presence of manic symptoms. Diagnosis of mania was made according to DSM-III-R criteria [36]. Cases were defined as secondary mania if there was no clear history of mood disorder, and no family history of mood disorder. Patients with confusional symptoms (acute brain syndrome) were excluded. Cases were interviewed by the treating psychiatry registrar and psychiatrist to obtain information regarding present and past psychiatric history and family history of psychiatric disorder. Collateral history was sought where possible. Cognitive testing was performed clinically by the registrar and consultant, and if abnormalities were noted, neuropsychological assessment was obtained. The psychiatry registrar and consultant determined treatment.
Methods
Prevalence of mania
During the 29 months from January 1993 to June 1995, 23 patients with mania were identified amongst the outpatient and inpatient population. Of these, four had a past history of mood disorder, two bipolar I disorder, and two major depressive disorder. No patients had a family history of mood disorder. Thus, during the study period 19 patients were identified as having secondary mania.
During the same time period 1553 HIV-positive patients attended Fairfield Hospital; 232 of these patients had AIDS. Thus, in this HIV-positive patient population the 29-month prevalence of secondary mania was 1.2%, and as 10 of these patients had AIDS at the time of their mania, the 29-month prevalence of secondary mania in patients with AIDS was 4.3%
Sociodemographic and HIV-related characteristics
Of the 19 patients, there were 17 males, and two females. The ages ranged from 22 to 59 years (mean=37.1 years). The HIV exposure category was homosexual or bisexual contact in 16 (84%), intravenous drug use in one (5%), and heterosexual contact in two (11%). This age range, male:female ratio and pattern of HIV exposure category is consistent with the overall pattern among HIV-positive patients attending Fairfield Hospital for treatment at that time. The CD4 count in these patients ranged from one to 220 cells/μL (mean=42.3). Among the 10 patients with AIDS, the commonest AIDS-defining illnesses recorded were HIV wasting syndrome (three), pneumocystis carinii pneumonia (two), and tuberculosis (two).
At the date of presentation with mania, seven patients were also experiencing symptoms associated with HIV infection: fevers (two), diarrhoea (one), inanition (one), cough (two), and headache (one). Among 18 patients admitted to hospital at the time of mania diagnosis, three had concurrent illness diagnosed: pulmonary tuberculosis (one), PCP(one), and probable mycobacterium avium bacteraemia (one). The remaining patient, treated as an outpatient, was physically well at the time.
Clinical features of secondary mania
All patients described a combination of mood disturbance and various symptoms typical of mania. Mood was most often described as a combination of elevated, irritable, and labile (10/19). Four patients were noted to have irritable and/or labile, but not elevated mood, and five had purely elevated mood. The DSM-III-R term, expansive, was applied by the assessing psychiatrist in only one case.
All of the 19 patients described a constellation of associated symptoms. All had at least three of the seven DSM-III-R symptoms of mania. Of note, each of the four patients whose mood was described only as irritable/labile had at least four DSM-III-R symptoms.
Decreased need for sleep (16/19), being more talkative than usual (15/19), flight of ideas (14/19), and an excessive involvement in pleasurable activities (14/19) were the most common symptoms. The average number of symptoms per patient was 4.6.
Seven of the 19 patients experienced delusions. Six were described as mood congruent, grandiose delusions. Three of these involved the theme of God favourably influencing the course of their AIDS, two involved themes of increased power and ability, and one had the theme of being able to read other people's minds. One patient experienced mood incongruent delusions, with bizarre and paranoid themes. One patient experienced hallucinations, described as the voice of God. There was no evidence of catatonia in any of the 19 patients.
The duration of symptoms was estimated from the patient's history, and corroborated by interview of family or carers when available. Symptoms had been present for 4–60 days (mean 20 days) prior to psychiatric assessment.
All patients were questioned about recent use of illicit and prescribed drugs. No patients reported the use of illicit drugs prior to the diagnosis of mania, and two reported ever having used illicit drugs. With respect to prescribed medications, two patients were taking low dose amitriptyline for the treatment of peripheral neuropathy (one for 6 months, the other for 12 months), and four were taking benzodiazepines on admission. Fifteen of the 19 patients were taking medication for various HIV-related illnesses. None was noted to be temporally related to the onset of mania.
Assessment of cognitive function during admission was difficult because features of the mania (distractibility and poor concentration) made testing unreliable; however, no patients had a prior diagnosis of AIDS dementia, and patients with confusional symptoms were excluded from the study. As a consequence, neuropsychological testing was not routinely performed.
Neuroradiological findings in secondary mania
All 19 patients had neuroradiological imaging performed to exclude focal cerebral pathology. Six had a computersied tomography (CT) scan, one magnetic resonance imaging (MRI) scan, and 12 had both a CT and an MRI scan.
Ten patients (53%) had an abnormality reported.
Computerised tomography scans
Eighteen of the 19 patients had a CTscan. Fourteen were reported as normal, three were reported as showing evidence of mild atrophy, and one showed a structural abnormality which was reported as ‘consistent with porencephalic cysts’.
Magnetic resonance imaging scans
Thirteen of the 19 patients had MRI scans. Six were reported as normal, and seven as abnormal. Of the seven abnormal MRI scans, four were reported as showing mild atrophy, and all seven were reported as showing small foci of high T2 signal intensity. These foci were all in the white matter bilaterally, usually in periventricular areas. In all cases these high T2 signal foci were reported to be of doubtful clinical significance.
Comparisons between computerised tomography and magnetic resonance imaging scans
Of the 12 patients who had both CT and MRI scans, five were reported as normal for both scans. Of the remaining seven, one had both MRI and CT reported as showing atrophy, one had a CT which showed atrophy but a normal MRI, and five had MRI's which showed high T2 signal foci (two with atrophy, three without) and normal CT scans. Hence, there were three cases in which atrophy was reported on one scan but not the other.
Significance of the scans
The scans were reported as being of doubtful clinical significance in all cases. Only one scan showed a structural abnormality (the porencephalic cysts) and this too was judged to be of no significance to the mania.
Treatment
The majority of the 19 patients (18/19) required inpatient care at Fairfield Hospital. Eleven were admitted for treatment of their mania, the remaining seven were admitted for other medical reasons. These included fever for investigation (two cases), treatment for pneumocystis carinii pneumonia, investigation of neurological symptoms (headaches), initiation of treatment for chronic diarrhoea, tuberculosis and respite care.
On admission, all patients were thoroughly assessed to exclude any treatable cause of the mania. All had past and recent serology for Treponema pallidum, Toxoplasma gondii, and Cryptococcus neoformans reviewed, thyroid function tests performed, T cell subsets measured and CT or MRI examination of the brain. Some, but not all had lumbar puncture to exclude meningeal disease. Although seven patients were admitted for management of comorbid physical complications of HIV disease, these illnesses and symptoms did not appear to be related to the mania. In particular, neurological examination was normal in all patients. No abnormalities were detected in those patients who had a lumbar puncture performed to enable microscopy and culture of the cerebrospinal fluid.
All 18 inpatients were managed by the C-L psychiatry service on the general medical ward at Fairfield Hospital. Four patients were admitted as involuntary patients under the Victorian Mental Health Act (1986). Three of the four patients admitted involuntarily were among the group with delusions, the fourth involuntary patient did not have psychotic symptoms, but required almost continual supervision because of behavioural disturbance. No patients required transfer to a secure psychiatric facility and none absconded.
The duration of admission ranged from 4 to 44 days (mean 18 days). The seven patients who were initially admitted for medical reasons stayed longer in hospital than those admitted for treatment of mania alone (mean of 24 vs 14 days).
All 19 patients were treated with neuroleptics. A mood stabiliser was prescribed in only one patient, and 12 received benzodiazepines.
Haloperidol was the most frequently used medication (17/19). One patient was treated with chlorpromazine, and one with trifluoperazine. All patients were started on a low dose of neuroleptics, usually 3–5 mg at night (in haloperidol-equivalents), and the dose was adjusted according to response and side effects. The single patient treated as an outpatient was prescribed trifluoperazine 10 mg.
Sixteen patients were discharged on neuroleptics. The discharge dosage in haloperidol-equivalents ranged from 1.5 to 50 mg/day (average 10.6 mg/day). The patient who received 50 mg was a clear exception to the majority. This patient was behaviourally disturbed, and failed to improve in mood or behaviour until the dose reached 60 mg. The dose was then reduced due to excessive sedation. There were no extrapyramidal side effects noted in this patient at any stage. This patient was one of the four treated involuntarily under the State Mental Health Act. Excluding this patient as an exceptional case, the discharge dosage in the remaining 15 patients ranged from 1.5 to 20 mg/day (average 7.9 mg/day).
One patient was treated with lithium carbonate at a dose of 750 mg b.d. (lithium level 0.8). This medication had been commenced at another hospital prior to transfer to Fairfield Hospital. In addition to lithium carbonate, the patient was prescribed chlorpromazine 600 mg and diazepam 10 mg. This patient was treated involuntarily under the State Mental Health Act and stayed longer than the average (38 days).
Twelve patients received benzodiazepines. For the majority (n=10), benzodiazepines were prescribed for night-time sedation: two were prescribed benzodiazepines for treatment of day-time agitation. The benzodiazepines prescribed were temazepam (n=5, mean dose=20 mg), flunitrazepam (n=2, mean dose=20 mg), diazepam (n=3, mean dose=21 mg) and clonazepam (n=2, mean dose=1.5 mg).
Side effects
Medication was generally well tolerated. Seven of the 19 patients developed extrapyramidal side effects (EPSE). The dosage of haloperidol in these patients ranged from 2.5 to 20 mg. All improved with a combination of lowering the dose of haloperidol (where appropriate) and the introduction of benztropine. The dose of benztropine ranged from 0.5 to 2.0 mg (mean 1.6 mg).
Three of the 19 patients developed an acute brain syndrome (ABS). The first of these patients was on haloperidol 10 mg and cogentin 2 mg as well as a range of HIV medications for treatment of mycobacterium avium complex (MAC) and candidiasis. This patient had advanced disease with a CD4 count of<10 cells/μL. Cessation of haloperidol and cogentin led to a resolution of the acute brain syndrome.
The second patient who developed an acute brain syndrome did so after cogentin 2 mg was added to haloperidol 7.5 mg for treatment of mild parkinsonian features. Cessation of cogentin lead to recovery within 2 days. This patient was relatively well medically, with no opportunistic conditions and no other medications.
The third patient with an acute brain syndrome had also been on multiple medications and had advanced HIV disease (CD4<10 cells/μL). The ABS developed 10 days after haloperidol 5 mg and temazepam 20 mg were commenced, and was associated with the development of PCP. This patient died as a result of PCPwithin 3 days of the diagnosis of the acute brain syndrome.
Outcome
With the exception of the patient who died, all improved and were discharged to outpatient care.
Follow-up did not always occur at Fairfield Hospital. Prior to May 1996, when the hospital closed, 15 of the 19 patients had been seen by the HIV psychiatry service at the hospital for follow-up of their manic episode. The duration of follow-up in these 15 patients ranged from one to 26 months.
Two of the 15 patients developed a further episode of mania, one 6 months after discharge, the other 2 months after discharge. Both responded to recommencement of neuroleptics without the need for readmission.
Two of the 15 patients experienced depressive episodes during the follow-up period, both 3 months after discharge. Both patients had ceased their neuroleptics by this time, and were treated with a combination of psychotherapy and fluoxetine with good response.
Cognitive testing was repeated on recovery from the manic episode, and during the follow-up period. If abnormalities were noted, formal neuropsychological testing was obtained. Five of the 15 patients followed up subsequently developed cognitive impairment. In two of these cases, the cognitive impairment was diagnosed within 3 months of discharge, and further follow-up revealed a clear deterioration leading to a diagnosis of AIDS-related dementia. One of these patients died 9 months after the index manic episode, the other after 11 months.
In the remaining three patients, the cognitive impairment was noted on neuropsychological testing, but a diagnosis of AIDS-related dementia was not made. Two of these patients died before a formal diagnosis of AIDS-related dementia could be established clinically or by further neuropsychological testing. The third patient continued attending outpatients with no evidence of further cognitive decline noted.
Of note, the five patients who were found clinically to have cognitive impairment all had normal CT scans. Three of the five had MRI scans, and only one was abnormal, showing high T2 signal foci.
Discussion
This pilot study provides descriptive data regarding the clinical features, neuroradiological features and treatment outcome in patients with secondary mania occurring in association with HIV infection. For this study, secondary mania was defined as mania occurring in the absence of a past or family history of mood disorder. This approach is limited by the patients recall, and the variable availability of collateral history. Krauthammer and Klerman suggested two other criteria: a close temporal proximity of the organic insult to the subsequent mania, and late age of onset. The high rate of HIV infection in a relatively young population and the prolonged course of HIV infection meant these criteria could not be readily applied in this medically ill group of patients (Krauthammer and Klerman's original work was based on patients with an acute organic insult usually occurring in older patient groups, e.g. those with a cerebrovascular accident).
Our estimate of the 29-month prevalence of secondary mania, based on hospital records of all patients attending over the study period, was 1.2% in patients with HIV infection, and 4.3% in patients with AIDS, considerably higher than the lifetime prevalence of bipolar disorder of 0.8% [37]. These figures must be viewed with caution given the various biases that may effect the number of cases and noncases attending a service such as Fairfield Hospital. Furthermore, it is possible that mania may be present in patients in the service who were not referred; however, in general these biases would operate in such a way as to lead these figures to be an underestimate of the magnitude of the problem.
As in previous descriptions of HIV-positive patients with secondary mania, our cases occurred in individuals with advanced disease, as shown by the average CD4 count of 42.3 cells/μL, and 10 of the cases having previously diagnosed AIDS. Overall, the features of the mania in this group were similar to those seen in patients with mania due to bipolar affective disorder as described in standard diagnostic criteria. The mood of the patients was mostly described as elevated (15 out of 19 cases). In the remaining four cases it was described as just irritable and labile. In total, 11 patients were noted to be irritable. Of interest, Kieburtz et al. [4] noted irritability to be common, occurring in six of their eight cases, as did Lyketsos et al. [23] occurring in eight of their 11 late onset manic cases. Our study confirms irritability is a common feature of secondary mania in HIV infection, but in the absence of comparative data about mania in other settings, it is not clear whether this represents a true difference between primary and secondary mania.
Neuroradiological investigations detected a high rate of cerebral abnormalities (10/19 patients) in patients with advanced HIV disease, but of note none of the abnormalities reported was of clinical significance with respect to the mania. Unlike the report by Halman et al. [3], MRI abnormalities in this group of patients did not predict higher rates of treatment intolerance or a poor response to treatment.
Treatment of mania secondary to HIV/AIDS (or other causes) has not been studied systematically, although many of the case reports note that control of manic symptoms is often difficult. This was not our experience. While treatment in some patients was complicated by comorbid illness, the mania did not appear more difficult to control than mania occurring in patients with bipolar disorder.
A review by Evans et al. [5] suggests that standard treatment approaches to mania may not be appropriate for the medically ill. Our results suggest a good response to neuroleptics, with side effect rates (apart from the development of ABS) that are similar to medically well population groups (7/19 with EPSE, 1/19 with excessive sedation). This may be a result of the relatively low doses of neuroleptics prescribed (mean dose 10.6 mg/day).
Three of the 19 patients developed an acute brain syndrome during their treatment for mania, possibly reflecting the increased risks of prescribing psychotropics (in particular benztropine) in medically unwell patient groups
The management of this group of patients with mania was undertaken on a medical ward in response to the need for the availability of both specialist HIV medical and psychiatric services. This was possible because of the presence of a full-time psychiatric registrar, and nursing staff who were experienced in psychiatric care.
Unfortunately, long-term follow-up was interrupted by closure of the hospital in May 1996, and the transfer of patients to several treatment facilities. However, even in the short time follow-up was possible, four of the 15 patients had a subsequent mood disorder (two mania, two depression). The findings regarding cognitive impairment are also limited by the short time of follow-up, but even so confirm the findings from the literature that cognitive deterioration does not always accompany secondary mania in patients with HIV infection. Some, but not all, patients appear to develop cognitive impairment.
In addition to the problems preventing long-term follow-up, the major limitations of this study are its uncontrolled design (ideally our cases would be matched with patients with primary mania to determine differences), and the inherent difficulties in defining secondary mania in an illness such as HIV which has so many manifestations. To ensure there were no clear causes for the mania other than HIV, patients were assessed by medical and psychiatric staff for exclusion of other possible causes such as intracerebral pathology (including a lumbar puncture) and illicit drug use.
Conclusions
Mania occurring in advanced HIV disease is not uncommon, and appears to be more prevalent than would be expected from epidemiological data regarding mania occurring in bipolar affective disorder. The characteristics of the secondary mania associated with HIV disease appeared on face value to be similar to those of primary mania, as was the response to treatment. The prognosis appeared worse, with a mortality of one in 19 (5.2%), and five of the 15 followed in outpatients developing cognitive impairment.
With regard to the concept of secondary mania, it is not clear from our pilot study, or other reports in the literature, whether mania associated with HIV disease is truly secondary. It may be that HIV infection or AIDS provided a stressor that precipitated the onset of mania in an individual vulnerable to bipolar disorder. The long course of HIV disease means a clear temporal relationship between the putative organic aetiological agent and the onset of mania is difficult to establish. In addition, the age of onset of bipolar affective disorder coincides with the age range of those affected by HIVdisease. We attempted to delineate a cohort of secondary mania by excluding those with a past or family history of mood disorder. Clearly some of our cohort could have been presenting with a first episode of coincident BAD. Of note, four patients in our study did develop a subsequent episode of mood disorder.
Lyketsos et al. [23] provide indirect evidence of a link between HIV-related mania and organic insult in that they found an increased risk of HIV dementia and psychomotor slowing in their cohort of patients with late onset mania in HIV infection.
To establish a clear link, the ideal approach would be to seek evidence of a temporal relationship between the onset of HIV-related brain pathology and the onset of mania. A study of this sort would require large numbers of patients and invasive tests. Alternatively, a case control study of the effects of standard treatments for HIV (such as antiretrovirals) on the incidence of mania in HIV-infected patients would provide further indirect evidence of the relationship between HIV infection and mania.
Footnotes
Acknowledgements
The authors would like to thank the Infectious Diseases Physicians at Fairfield Hospital for their support in establishing and developing the Consultation-Liaison Psychiatry Service, the nursing staff who enabled the management of the patients in a general medical ward, the liaison psychiatry nurse, Margaret Grigg, and Dr John Lloyd for his neuropsychiatric evaluations.