Sustained Corticosteroid-Induced Mania and Psychosis Despite Cessation: A Case Study and Brief Literature Review
|Mary Gable1* and Dwayne Depry DO2|
|1Psychiatry Department, UCSF Fresno, USA|
|2Chief of Inpatient Psychiatry, VA Central California, USA|
|Corresponding Author :||Mary Gable|
Psychiatry Department, UCSF Fresno
155 North Fresno Street, Fresno CA 93701, USA
Tel: +1 559-499-6400
|Received June 04, 2015; Accepted June 18, 2015; Published June 20, 2015|
|Citation: Gable M, Dwayne Depry DO (2015) Sustained Corticosteroid-Induced Mania and Psychosis Despite Cessation: A Case Study and Brief Literature Review. J Clin Case Rep 5:546. doi:10.4172/2165-7920.1000546|
|Copyright: © 2015 Gable M, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.|
Visit for more related articles at Journal of Clinical Case Reports
View PDF Download PDF
Objective: Corticosteroids generally result in short-lasting neuropsychiatric symptoms following cessation, but
the following case highlights an unusually long-lasting course of symptoms in a patient following near immediate
cessation of medication, despite medication management and electroconvulsive therapy. The case presentation
will be followed by a discussion of the presentation, treatment, mechanism, and management of steroid induced
Methods: The patient was followed from symptom onset to resolution.
Results: The patient’s symptom course was unusually long and required a long course of multi-modal therapy.
Conclusions: Corticosteroids are commonly used medications both in a wide variety of medical settings,
and despite this, their neuropsychiatric effects are poorly understood. The affective and behavioral symptoms, in
particular mania and psychosis, can be unpredictable and challenging to treat as in our patient, who developed a
long lasting psychotic episode on high dose steroids, despite having tolerated them multiple times in the past and
whose very marked symptoms persisted, despite discontinuation and treatment for almost 6 months.
|Psychosis; Corticosteroid; Treatment; Duration; Risk|
|Corticosteroids are very commonly prescribed among both hospitalized and outpatient medical and surgical patients. While many of the common endocrinological and pathological side effects of corticosteroids are well understood, the incidence, presentation, and treatment of neuropsychiatric manifestations have not been commonly characterized and studied. Though most practitioners are familiar with the term “steroid-induced psychosis,” the affective and behavioral symptoms related to corticosteroid use are varied and can be unpredictable. In many cases, the appropriate treatment and its duration can be elusive. We present a case of corticosteroid induced mania and psychosis, highlighting a highly unusual course and resolution, as well as a prolonged duration despite an aggressive treatment course, including electroconvulsive therapy that has not been noted in medical literature based on our review. The case is followed by a discussion of incidence, risk factors, and treatment options.|
|A 58 year old male with a history of general anxiety disorder who was brought into the emergency room on an involuntary hold for unusual behavior and an inability to care for himself. He had been seen wandering around a store parking lot with delusional thinking and pressured speech. On exam, his thought process was tangential, confused and circumstantial. The patient denied auditory or visual hallucinations at the time, and the patient and his family attributed his condition on admission to his current steroid use. His urine toxicology screen was negative, and there was no history of prior drug or alcohol use. Mr. A. was alert and oriented to person, place, time, and situation. The patient had no other significant prior psychiatric history, other than a diagnosis of anxiety, including no previous history of reported mania, psychosis or bipolar disorder. An MRI of his head, blood count with the exception of platelet count, chemistry panel, thyroid studies, and ammonia level were normal.|
|The patient had a history of idiopathic thrombocytopenic purpura which had been diagnosed 9 years prior to this admission. He had had his first relapse two years after diagnosis and then again four years subsequent to that, and a third time about two years prior to this admission. He was on a 90 mg prednisone taper for his first two relapses, each just over two months in duration, and a 100 mg prednisone taper for the third, lasting about four weeks. He was on a taper for about four weeks just three months prior to this admission because of biopsy negative mediastinal lymphadenopathy which had subsequently resolved. The patient had never had a prior reaction to steroids. Then almost eight weeks prior to admission, he had a notable drop in platelets, and he was started on four day of 40 mg of dexamethasone bursts every three weeks. When he was on his third cycle, his behavior changed, and per his family, he was behaving in a bizarre manner and could not care for himself.|
|The patient was admitted to the medicine floor for low platelets, and his dexamethasone was stopped. He continued to be inappropriate and to demonstrate evidence of mania and psychotic symptoms. Psychiatry was consulted, and the patient was started low dose quetiapine. He was observed to be naked at night in his room, to be praying in unintelligible speech, and to have disorganized thoughts and manner of speaking.|
|After 10 days admission to medicine, the patient’s platelets were stabilized, and he was transferred to inpatient psychiatry for further stabilization of his symptoms which included minimal sleep. The patient was started on a dose of night time quetiapine which was up-titrated to 200 mg in the morning and 600 mg at bedtime. Sleep improved, and the patient became more directable. However, he continued to engage in hyper-religious behaviors such as “speaking in tongues” and chanting for hours. A short course of aripiperazole to which he had a dystonic reaction followed by 5 days of electroconvulsive therapy brought about no additional improvement. High-dose quetiapine was re-started, and the patient gradually improved over a total of almost 6 months.|
|Corticosteroids have been in use for over six decades to treat a wide variety of pathologies from asthma and allergies to autoimmune diseases and dermatologic conditions. Corticosteroid use is common, and up to 10% of inpatients receive steroids, while about 2 to 3% of the general population is on long-term glucocorticoid treatment . It is estimated that about 20% of patients who receive high dose corticosteroids, defined as greater than 40 mg of prednisone or its equivalent, will develop a psychiatric disorder like mania, psychosis, or depression severe enough that with condition will require pharmacotherapy [2,3]. In a large United Kingdom study that examined steroid use in over 350,000 patients, it was determined that suicides and suicide attempts were almost seven times as common in those who had been prescribed steroids for a condition when compared with those with the same condition who had not received the medication. Additionally, mania was over four times as common, and disorientation and confusion were noted at five times the rate. The overall incidence of neuropsychiatric effects of corticosteroids is not entirely known with reports ranging anywhere from 2 to 60%, and with a weighted average of 6% in metaanalyses. The majority of patients, 60 to 85%, will develop symptoms within the first week of treatment, and about 90% within six weeks of initiating treatment, with at least half noting symptoms within the first ten days as did our patient .|
|Manic episodes constitute the most common manifestations of corticosteroid-induced neuropsychiatric disorder (CIPD), accounting for about one-half of all psychiatric referrals and make up an estimated 40% or more of those with a neuropsychiatric manifestation. Psychotic symptoms are prominent among those with mania, seen in about 30 to 40% of those patients. The literature suggests that patients who suffer from steroid-induced mania are more likely to demonstrate psychotic features than are patients who have non-substance induced mania. Persecutory delusions, auditory hallucinations, and disorganized behaviors are prevalent psychotic symptoms among such patients. Patients usually present with prominent signs of mania including excitation, pressured speech, euphoria, hyperactivity, and irritability. In one study that reported on eighteen patients who were referred to for psychiatric consultation over a 10 year period, of the 2069 patients referred, 18 were noted to have met DSM IV criteria for CIPD and had no prior psychiatric history. The majority of those patients, fifteen of the eighteen in total, presented with mania, and three of those also had psychotic symptoms [5-8].|
|Even withdrawal of corticosteroids can initiate the commonly termed “withdrawal syndrome,” the latter resulting in anorexia, depression, fatigue, poor concentration, depersonalization, and psychosis with additional symptoms including lability, somatic concerns such as paresthesias and faintness, poor memory and focus, depersonalization, and in some case, suicide [8,9]. Such side effects of withdrawal are typically short-lived, but have been reported to last anywhere from two to eight weeks post-termination .|
|Prior psychiatric history, age, underlying disease, and gender have been postulated to serve as risk factors for the development of CIPD. However, studies have shown mixed results. Hall et al. found no correlation with a history of a previous psychiatric condition or CIPD, while another study by Wada et al. nine of eighteen patients had recurring CIPD, suggesting that treatment with corticosteroids had induced a mood disorder that was self-propagating With respect to age, there appears to be no association with CIPD. Neuropsychiatric manifestations also do not appear to be associated with a specific underlying disease either with the exception of systemic lupus erythematosus which increases the risk by a factor of two or more. Gender however does pose a risk. Even when controlling for the higher incidence of certain medical conditions requiring corticosteroid treatment in women, neuropsychiatric symptoms are up to six times more common in women when compared with men. However, as in our patient, previous tolerance of corticosteroids is not correlated with a reduced frequency of neuropsychiatric side effects on re-introduction of the medication [9,10].|
|There is also a dose dependent relationship that causes the development of neuropsychiatric symptoms with fewer than 2% of patients being affected at doses less than 40 mg per day of prednisone or its equivalent, while around 5% of those dosed between 40 and 80 mg will have psychiatric manifestations. Nearly 1 in 5 of those on 80 mg or more per day will be affected. Mr. A was taking the equivalent of just over 260 mg of prednisone daily. While there are case reports of psychiatric manifestations in patients on even very low doses, there does appear to be a dose dependent relationship, and the average dose of those who present with psychosis is around 60 milligrams [9-12].|
|There are no FDA-approved medications to manage steroid induced neuropsychiatric symptoms. Therefore, treatment is based on anecdotal evidence and case reports as well as a few small, scattered case series. However, the first step in managing any steroid-induced psychiatric symptom is to reduce the dosage of the medication to that of less than 40 mg per day of prednisone or its equivalent, and ideally, if possible, to stop the drug. About three-fourths of psychiatric symptoms will remit on discontinuation alone [11,13]. Lewis and Smith  found that approximately 90% of patients treated with a taper only achieved recovery. Eighty-four percent of those treated with neuroleptics recovered. All of those treated with electroconvulsive therapy (ECT) or taper in addition to either lithium or neuroleptics had symptom resolution in case reports and small studies, but our patient appears to stand out in terms of his long duration of symptoms with minimal response to ECT.|
|For depressive symptoms, tri-cyclics are generally avoided as their anti-cholinergic effects can exacerbate or worsen delirium states which can often be coexistent. Serotonin reuptake inhibitors and serotonin norepinephrine reuptake inhibitors have been reported to have been successfully used in case reports. Lithium has also been used to treat depression alone.|
|Carabamazapine, lamotrigine, and valproate are effective in treating corticosteroid induced psychosis and/or mania. In multiple studies [4-6], phenothiazines have proven effective. Haloperidol can be used to control the majority of psychotic reactions to corticosteroids. Olanzapine and risperidone have been used to treat psychiatric reactions including mania and depression, and in one report, risperidone proved effective in treating a patient in whom a reduction in corticosteroids was not possible in the short term. In a case study of an adolescent steroid induced mania and psychosis showed Seroquel to be useful in resolving symptoms over a matter of weeks. Brown et al.,  found that olanzapine was effective in 92% of patients who suffered from mania or mixed symptoms at doses ranging from 2.5 to 20 milligrams per day, resolving symptoms over no more than a five week period.|
|In half of patients, termination of the steroid will lead to symptom resolution within about two weeks, while in 90% of patients, symptoms will cease within 6 weeks of cessation. Complete recovery is ultimately expected in nearly all patients, though this can be out as far as six months from treatment cessation. It is not known whether the addition of anti-psychotics or other medications results in or hastens symptom resolution, but it is likely that, at least, its use controls symptoms [7,8,12].|
|There may be useful symptom prevention strategies when giving patients steroids, though none would have directly applied to our patient. Dosing plays a significant role in the development of neuropsychiatric symptoms, thus minimizing the dose is a primary preventive strategy with the goal of maintaining a 40 mg daily dose of prednisone or its equivalent. Patients who are at higher risk for psychiatric complications, specifically those with a damaged blood brain barrier and those with hypoalbuminemia, which increases the risk by a factor of two , should be carefully considered for steroid therapy. Avoiding other drugs that may increase circulating levels of corticosteroids is also important. For example, Clarithromycin, a P450 CYP 3A4 inhibitor that prevents breakdown of prednisolone, prednisone’s biologically active metabolite, has been shown to incite mania . There are small trials that have supported the use of prophylaxis, particularly in those with a prior history of steroid induced psychosis or mania using such agents as olanzapine and lithium . Dosing regimen, such as pulsed dosing, daily divided dosing, and alternate day dosing, does not seem to reliably prevent the development of neuropsychiatric symptoms, and in fact the latter has been associated with rapid cycling of moods .|
|Corticosteroid induced neuropsychiatric symptoms are not uncommon. Despite this, there are few sizable studies on their presentation and management. Medication cessation or dosage should always be the first line of treatment, and the addition of anti-psychotics is supported by the literature. Though further research is needed to determine whether such medication hastens or causes symptom resolution, it would appear to contain symptoms, though it did not have an impressive effect in the patient described herein. ECT has, in small reports, been shown to be highly effective in treating steroid induced symptoms, though, again, this was not the case in this patient. The prognosis for patients in general is excellent with expected complete resolution in nearly all of those affected, though this may take several weeks or even many months.|
- World Cardiology Conference
17-18 September,2018 Tin Shui Wai, Hong Kong
- Clinical and Experimental Dermatology
November 02 - 04, 2018 San Francisco, USA
- Total views: 11702
- [From(publication date):
June-2015 - Mar 14, 2018]
- Breakdown by view type
- HTML page views : 7927
- PDF downloads : 3775
Psychiatric complications of treatment with corticosteroids: Review with case report
Heather A. Kenna MA,Corresponding author
- Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, USA
Amy W. Poon MD,
- Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, USA
C. Paula de los Angeles BA,
- Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, USA
Lorrin M. Koran MD
- Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, USA
None of the authors have any conflicts of interest to report.
This manuscript was not sponsored by any grant funding support.
Heather Kenna, MA, Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, 401 Quarry Road, Room 2360, Stanford, CA 94305-5723, USA. Email: email@example.com
Corticosteroids are widely used in modern medicine but can result in troubling psychiatric side-effects. Physicians and other medical professionals should be aware of the potential for these side-effects, possible means of prevention, and efficacious treatments. Herein, we review adult case report data published during the past quarter-century on adverse corticosteroid-induced psychiatric effects, and present a case of corticosteroid-induced psychotic depression. PubMed and PsychLit databases were searched using the terms ‘corticosteroids’, ‘steroids’, and the generic names of corticosteroid medications with terms for psychiatric symptoms or syndromes, including psychosis, mania, hypomania, depression, apathy, anxiety, panic, depersonalization, delirium, confusion, hallucinations, delusions, paranoia, cognitive impairment and dementia. Fifty-five cases and a number of clinical trials investigating the incidence and treatment of these psychiatric symptoms and syndromes were identified. Data on incidence, drug dose, risk factors, course of illness and treatment (when present) were tabulated. We conclude that the cumulative data indicate that psychiatric complications of corticosteroid treatment are not rare and range from clinically significant anxiety and insomnia, to severe mood and psychotic disorders, delirium and dementia. While tapering or discontinuation of the corticosteroid treatment may remedy these adverse side-effects, psychotropic medications are often required because of the medical necessity of the corticosteroid or the severity of the psychiatric symptom. Further studies are needed to better understand the deleterious psychiatric effects associated with corticosteroids.
CORTICOSTEROIDS ARE A widely used and highly effective treatment for a number of conditions, including immunologic and inflammatory disorders,1 systemic lupus erythematosus (SLE) and systemic vasculitis,2,3 asthma and chronic obstructive pulmonary disease,4,5 cancer,6 acute and chronic back pain,7,8 and in the prevention of postoperative swelling in head and neck surgery.9–11 Unfortunately, troubling psychiatric side-effects are sometimes seen in patients treated with corticosteroids.12–14 The pathophysiology of negative reactions to corticosteroid administration or withdrawal is not well understood, but clinical practice may benefit from greater awareness of these potential adverse events and of methods to possibly prevent and to treat them.
CORTICOSTEROID-INDUCED SYMPTOMS AND SYNDROMES
The psychiatric signs, symptoms and syndromes associated with corticosteroid treatment include (DSM-IV substance-induced) mood disorders (hypomania, mania, mixed states, depression), anxiety and panic disorder,15,16 delirium, suicidal thinking and behavior in the context of affective syndromes or delirium,12,17–20 aggressive behavior21 (including attempted murder22), insomnia and agitation with clear consciousness, depersonalization; and, isolated cognitive impairments (impaired attention, concentration, memory and word-finding difficulties).14,17,23–26 In addition, the existence of a reversible, corticosteroid-induced dementia has been confirmed.27–30 Rarely, corticosteroids have been abused for their euphoria-producing effects, producing drug dependency.31–33 The term ‘psychosis’ has been variably applied to many of these clinical presentations, without distinguishing, for example, mania or psychotic depression from delirium.
Corticosteroid-induced cognitive deficits in the absence of psychotic symptoms commonly involve declarative or verbal memory.34–40 Similar reversible deficits in declarative memory have been reported in Cushing's disease and are greater in more severe cases,41–43 suggesting that excess endogenous and exogenous corticosteroids produce similar cognitive impairment. Reversible cognitive deficits and mood symptoms have been reported in healthy control subjects after administration of prednisone,40 dexamethasone,37,39 and cortisol.38 By contrast, a small (n = 10) double-blind, placebo-controlled, crossover-design study utilizing prednisone 0.5 mg/kg/day in SLE patients with mild symptoms who had not received corticosteroids for at least 6 months demonstrated beneficial effects on cognition, mood and SLE symptoms.44
In older patients, corticosteroid-induced dementia has been misdiagnosed as early Alzheimer's disease and can occur in patients who have not experienced a steroid psychosis and are free of mood symptoms.45 Deficits occur in memory, attention, concentration, and mental speed and efficiency, and in severe cases, formal IQ is substantially reduced.28,45 Occupational performance is diminished, but the patients do not appear ‘manifestly demented, amnestic, or disoriented, or . . . toxic or intoxicated’ (p. 372).45
INCIDENCE OF PSYCHIATRIC SYMPTOMS AND SYNDROMES
Almost 40 years ago, The Boston Collaborative Drug Surveillance Program46 reported ‘psychiatric reactions’ in 1.3% of 463 patients treated with 40 mg/day or less of prednisone, 4.6% of 175 patients dosed with 41–80 mg/day, and 18.4% of 38 patients receiving doses higher than 80 mg/day. A decade later, Lewis and Smith12 reported a weighted average 5.7% incidence of severe psychiatric symptoms across 13 studies involving 2555 patients treated with corticosteroids. A subsequent review by Steiffel et al.6 of major psychiatric symptoms in cancer patients treated with high-dose corticosteroids noted a 5–10% incidence. In contrast to patients with other medical conditions, cancer patients seemed more likely to develop delirium rather than affective syndromes, and these authors postulated that this difference may arise from the patients' frequent medical complications, from the co-administration of narcotics, and from the neurological effects of the cancers themselves. Nishimura et al.,47 reviewing 139 treatment episodes in 135 patients with SLE but without current overt central nervous system manifestations, observed 14 cases (10.1%) of new-onset DSM-IV disorders, primarily manic episodes (n = 9, 6.5%).
In the older literature, psychotic symptoms were reported in 58% of the 55 cases reviewed by Ling et al.14 Nearly 72% of their 55 cases included mood symptoms. In the review by Lewis and Smith,12‘disturbances in reality testing’ were reported in 71% of the 79 cases (which included 44 of the 55 cases from the review of Ling et al.14), but only 14% (11/79) had ‘a psychotic disorder without evidence of significant mood changes or features of a delirium.’ Depression was present in 32/79 (40.5%), mania in 22/79 (27.8%), a mixed state in 6/79 (7.6%) and delirium in 8/79 (10%) of the sample. A review by Sirois48 reporting on steroid ‘psychosis’ published between 1970 and 1983, before publication of the current edition of the American Psychiatric Association's DSM-IV-TR, found a syndromal breakdown of 35% mania, 28% depression, 12% mania and depression, 13% delirium, and 11% psychosis.
A number of more recent publications support the conclusion that symptoms of hypomania or mania are the most common psychiatric adverse effect of corticosteroid treatment.12,13,18,20,47,49 Some recent studies have suggested, however, that the risk of depression increases with prolonged or chronic exposure.48,50–52 Patients who experience corticosteroid-induced depression during one treatment course may experience drug-induced mania in a subsequent course, and vice versa.53
Appenzeller et al.54 report episodes of acute psychosis meeting DSM-IV criteria in 17.1% (89/520) of SLE patients followed for periods of from 4 to 8.8 years. Corticosteroid-induced psychoses accounted for 31.5% (28/89) of the cases of psychosis (a 5.4% incidence in the entire sample), of whom 10 (35.7%) had more than one psychotic episode.
Gift et al.55 found significantly greater self-reported depression scores in 20 patients with chronic obstructive pulmonary disease receiving 20–40 mg/day of prednisone for 10–14 days than in 20 not receiving corticosteroids, but did not include any measure of hypomanic/manic symptoms. Swinburn et al.,56 in a study of 20 similar patients given 30 mg/day of prednisolone, found only a small reduction in anxiety and depression after 3 days of prednisolone (before lung function improved) and no significant mood change after 2 weeks of treatment.
ONSET OF PSYCHIATRIC SYMPTOMS
Early studies12,14,26 suggested and later studies13,39,57 confirmed that the psychiatric side-effects of corticosteroid treatment have a rapid onset. Lewis and Smith12 reported a median time to onset of 11.5 days; 39% of cases had onset during the first week and 62% within 2 weeks. Hall et al.26 noted that 86% of patients with psychiatric side-effects developed these symptoms within 1 week of starting treatment. Ling et al.14 reported that the psychiatric sequelae of corticosteroid treatment generally occurred within 2 weeks. In a study of 50 ophthalmic patients receiving high-dose corticosteroids, Naber et al.13 reported development of psychiatric symptoms within 3 days of initiating corticosteroid treatment (when present). Another study by Nishimura et al.47 found a mean of 12.5 days for onset of symptoms in their 14 cases, noting a range of 2–28 days after starting treatment. Lastly, a study of healthy subjects by Wolkowitz et al.39 reported onset of psychiatric sequelae within 5 days of corticosteroid administration.
The corticosteroids dexamethasone and betamethasone have half-lives of 36–54 h.58 As a result, they may accumulate and induce psychiatric symptoms that begin after the last dose has been given.9,10
CORRELATION BETWEEN DOSE AND OCCURRENCE OF PSYCHIATRIC SYMPTOMS
The early study of the Boston Collaborative Drug Surveillance Program demonstrated the striking dose–response correlation of corticosteroids. Similarly, Chan et al.59 reported psychosis in 8% of patients receiving prednisone 90 mg/day compared to 3% of patients receiving 30 mg/day. The more recent literature confirms that the likelihood of inducing psychiatric symptoms follows a dose–response correlation. Nishimura et al.47 noted that all 20 SLE patients who developed psychiatric symptoms (primarily hypomania, mixed states or depression) were receiving at least 40 mg/day of prednisolone. An additional 97 patients receiving these doses did not develop psychiatric symptoms. Wada et al.,18 describing 18 patients who developed mood disorders or psychosis after receiving 30–60 mg/day of prednisone-equivalent, also reported a strong association with dose, although they noted that some patients had a recurrence of depression or mania related to psychosocial stressors rather than to dose changes or to resumption of corticosteroids. Appenzeller et al.54 reported that all patients with corticosteroid-induced psychosis were taking prednisone 0.75–1.0 mg/kg/day, which translates into total doses as high or higher than those just noted.
Olsen et al.60 found a significant correlation between mood lability and prednisone dose in mg/kg during a 6-week taper from 40 mg/d to zero in 32 patients with alopecia areata. Naber et al.13 used the Profile of Mood States scale, a European psychiatric symptom scale developed by the Association of Methodology and Documentation in Psychiatry, a semi-structured interview and a battery of neuropsychological tests to study the psychological and cognitive effects of methylprednisone or fluocortolone (50–300 mg/day to start, tapered to 18–100 mg/day by day 8) in 50 ophthalmologic patients, all of whom were initially free of psychiatric disorders. Although 36% (18/50) developed DSM-III-R mania (n = 13) or depression (n = 5) during treatment, the authors found no correlation between the daily dose of steroid and the daily ratings of mood symptoms. No patient developed psychosis, delirium or severe cognitive deficits.
In almost all of the cases of corticosteroid-induced dementia located in the present search of the literature, the corticosteroid dose has been at least 60 mg/day of prednisone-equivalent. Varney et al.45 reported one case in which dementia was documented 2 weeks after the prednisone dose was reduced to 20 mg/day after a 4-month course of 100 mg/day; irregular improvement occurred over the following 24 months, while prednisone was continued at 20 mg/day.
OTHER RISK FACTORS
Other than dose, no strong predictors of risk have emerged. However, psychiatric risk may be increased by drugs that increase circulating levels of corticosteroids. Clarithromycin, for example, is an inhibitor of the cytochrome P450 enzyme (CYP) 3A4 that metabolizes prednisone's biologically active metabolite, prednisolone. Finkenbine and Gill21 reported a case of mania induced by adding clarithromycin to prednisone; the mania resolved over 5 days when both drugs were stopped. Finkenbine and Frye21 report a case of psychotic paranoia that required the addition of olanzapine, plus a prednisone taper and the discontinuation of clarithromycin, which was followed over the next 6 days by a clearing of the paranoia.
Nishimura et al.,47 studying 135 patients with systemic SLE, found cerebral spinal fluid/serum albumin ratio (a marker of blood–brain barrier damage) to be a significant risk factor for corticosteroid-induced psychiatric disorder (odds ratio 33.3). Chau and Mok,61 studying 92 SLE patients, of whom 5% experienced a corticosteroid-induced psychosis or mania, found that hypoalbuminemia distinguished those who suffered this side-effect from those who did not. Appenzeller et al.54 reported that after multiple regression analysis, hypoalbuminuria (odds ratio 2.2) was the only variable significantly associated with corticosteroid-induced psychosis in their series of SLE patients.
Some earlier studies,12,18,62 but not all,13,63 reported a higher prevalence of steroid-induced psychiatric problems in women. In part, the higher prevalence may have reflected their greater propensity to seek medical care, more common experience of certain psychiatric disorders, such as major depression and most anxiety disorders, or the higher prevalence of certain medical disorders, such as SLE in women; the female preponderance persisted in one study (2) even after cases of SLE were excluded.
A previous history of psychiatric disorder does not seem to increase the risk of an adverse psychiatric reaction.6,64 Early studies reported that prior episodes of steroid-induced psychiatric symptoms may or may not be followed by recurrence during future treatment courses12,26,65 and that risk does not appear to be associated with a particular age group.12,14 Nothing in the modern literature contradicts these findings.
The pathophysiological mechanisms giving rise to the psychiatric symptoms associated with corticosteroid treatment remain unclear. Speculations regarding these mechanisms are discussed elsewhere5,20,39,40,47,66 and include corticosteroid effects on dopaminergic and cholinergic systems,67,68 decreases in serotonin release,69 and toxic effects on hippocampal neurons39 or on other brain regions.28
RESULTS OF OUR REVIEW OF CASES
The following review of the adult case report data on corticosteroid-induced psychiatric side-effects targets those published since the last major review of this kind in 1983 by Lewis and Smith,1 with the goal of ascertaining whether recent experience with corticosteroids has produced new conclusions regarding the clinical pictures and their management. We searched the PubMed and PsychLit databases by combining the search terms ‘steroids’, ‘corticosteroids’ and the generic names of corticosteroid steroid medications with terms for psychiatric symptoms or syndromes including psychosis, mania, hypomania, depression, apathy, anxiety, panic, depersonalization, delirium, confusion, hallucinations, delusions, paranoia, cognitive impairment and dementia. Our search produced 55 cases along with a number of trials investigating the incidence and treatment of these conditions. Our review focuses on the nature of these adverse events, their incidence, correlation to drug dose, risk factors, course and treatment; we also summarize the findings from earlier studies and reviews. A caveat is that case reports are likely to be influenced by publications bias – more dramatic or consequential symptoms or syndromes are more likely to be written up and accepted for publication. Thus, the analyses we present are unlikely to represent accurately the milder symptoms that may occur or the true population incidence rates of more serious reactions. Still, our analyses may provide value to the clinician treating such patients.
Our search produced 55 cases with the following syndromes: hypomania/mania; depression; delirium; subsyndromal symptoms, such as hallucinations and agitation/anxiety, and panic disorder. Of these cases, 34 (61.8%) were psychotic, that is, had hallucinations and/or delusions coupled with impaired reality testing or lack of insight. Suicidal ideation was present in 22 cases (40%), of whom half were psychotic and half not; one patient committed suicide.53 In addition, the search produced three cases of steroid-induced reversible dementia confirming the 1984 observations of Varney et al.27–30,45 and six cases of psychoses apparently induced by rapid steroid discontinuation.70–76 The dementia cases were atypical of the corticosteroid cases we identified in that only two patients were younger than age 50 (ages 25 and 44 years), symptom onset was often not reported until after months of corticosteroid treatment, and recovery, not always complete, often took more than 6 months after corticosteroid discontinuation.
We analyzed the 55 non-dementia cases of psychiatric syndromes induced by corticosteroid administration to characterize the patients, the drugs involved, the psychiatric symptoms and treatments. A summary of these cases is provided in Table 1. Although in some cases, particularly cases associated with SLE, contributions of the underlying disease or of psychosocial stressors could not be ruled out, the results of our analysis largely resemble those of earlier case series analyses.12–14
|Ferris et al. 20039||Female||62||Undergoing parotidectomy||160 mg (p.d.e.)|
|Ismail et al. 20022||Female||64||Asthma||30 mg|
|Ito et al. 200377||Female||37||Acute myeloid leukemia||50 mg|
|Jenkins et al. 200066||Female||53||Meningioma||106.7 mg (p.d.e.)|
|Terao et al. 199478||Male||42||Systemic lupus erythematosus||40 mg|
|Terao et al. 199478||Male||18||Systemic lupus erythematosus||15 mg|
|Terao et al. 199779||Female||24||Systemic lupus erythematosus||50 mg|
|Terao et al. 199779||Female||66||Polyarteritis nodosa||60 mg|
|Wada et al. 200053||Female||31||Systemic lupus erythematosus||30 mg|
|Wyszynski et al. 199380||Female||46||Sjögren's syndrome||156.3 mg (p.d.e.)|
|Yoshimura et al. 200181||Male||70||Pneumonia||100 mg|
|Yoshimura et al. 200181||Female||50||Pneumonia||80 mg|
|Benyamin et al. 200882||Male||67||Chronic neck pain||100 mg (p.d.e.)|
|Bloch et al. 199483||Male||26||Multiple sclerosis||30 mg|
|Brown et al. 199984||Female||21||Asthma||30|
|Brown et al. 200149||Female||43||Behçet's disease||30 mg|
|Cerullo et al. 200685||Male||69||Unidentified cancer||Not reported|
|d'Orban 198986||Male||26||Undergoing maxillary osteotomy||60 mg (p.d.e.)|
|Finkenbine et al. 199787||Female||30||Sinusitis||30 mg|
|Franco et al. 200088||Male||59||Post-stroke cardiac transplantation||20 mg|
|Ginsberg et al. 200189||Female||59||Asthma||35 mg|
|Hong et al. 200690||Female||48||Sheehan's syndrome||15 mg|
|Johnson et al. 199676||Female||36||Asthma, pregnancy||20 mg|
|Kato et al. 200591||Female||46||Systemic lupus erythematosus||40 mg|
|Lopez-Medrano et al. 20023||Female||20||Systemic lupus erythematosus||60 mg|
|Lopez-Medrano et al. 20023||Female||21||Systemic lupus erythematosus||30 mg|
|Lundberg et al. 200092||Male||32||Hodgkin's lymphoma||100 mg|
|Muzyk et al. 201093||Female||31||Systemic lupus erythematosus||60 mg|
|Preda et al. 199994||Female||41||Nephritis||40 mg|
|Siddiqui et al. 200595||Male||52||Liver transplant||225 mg (p.d.e.)|
|Viswanathan et al. 198996||Male||38||Post-renal transplant||15 mg|
|Wada et al. 200053||Female||40||Dermatomyositis||60 mg|
|Wada et al. 200053||Female||19||Minimal-change nephrotic syndrome||30 mg (p.d.e.)|
|Wada et al. 200053||Female||21||Dermatomyositis||50 mg|
|Wada et al. 200053||Female||23||Ulcerative colitis||26.67 mg (p.d.e.)|
|Wada et al. 200053||Female||47||Systemic lupus erythematosus||Not reported|
|Wada et al. 200053||Male||68||Intractable nephrotic syndrome||15 mg|
|Wada et al. 200053||Male||53||Systemic lupus erythematosus||40 mg|
|Wada et al. 200053||Male||42||Kidney transplant rejection||666.7 mg (p.d.e.)|
|Ahmad et al. 19994||Female||55||Chronic obstructive pulmonary disease||60 mg|
|Artukoglu et al. 200797||Female||32||Biopsy for humeral mass operation||53.33 mg (p.d.e.)|
|Benazzi et al. 199798||Female||70||Systemic lupus erythematosus||5 mg (p.d.e.)|
|Galen et al. 199710||Male||26||Facial osteotomies||666.7 mg (p.d.e.)|
|Jenkins et al. 200066||Male||93||Renal cell carcinoma||133.3 mg (p.d.e.)|
|Koh et al. 200299||Male||40||ER-breathing difficulty||82 mg|
|Mada et al. 2009100||Female||72||Adrenal insufficiency||12.5 mg (p.d.e.)|
|Okishiro et al. 2009101||Male||67||Pulmonary emphysema||13.33 mg (p.d.e.)|
|Silva et al. 1995102||Male||45||Dental operation||100 mg (p.d.e.)|
|Stoudemire et al. 199631||Female||40||Chronic obstructive pulmonary disease||40–100 mg|
|Charbonneau et al. 199716||Female||31||Erythema multiforme||5 mg|
|Raskin et al. 198415||Female||35||Hirsutism||Not reported|
|Daragon et al. 1997103||Male||72||Rheumatoid arthritis||74 mg (p.d.e.)|
|Gallerani et al. 2008104||Male||64||Hypertension||40 mg (p.d.e.)|
|Finkenbine et al. 199821||Male||50||Emphysema||20 mg|
Patients' ages ranged from 18 to 93 years, with a mean of 44.5 ± 17.7 years. As in some earlier reviews,12,62 but not all,13,63 more cases involved women (34/55, 61.8%) than men. Symptoms began a mean of 12.2 ± 13.7 days after starting corticosteroids (n = 50), and within 1 week in 60% (30/50), but onset ranged from 1 to 60 days after starting the drug.
Hypomania or mania was the most common presentation, present in 54.5% (30/55) of cases. Clinical depression was present in 23.6% (13/55), of whom two patients were also delirious, and delirium was present in 20% (11/55). Suicidal ideation was reported in 36.4% (20/55), of whom a little less than half (9/20) were also psychotic. ‘Psychosis’, or a psychotic mania, psychotic depression or delirium was reported in 61.8% (34/55), quite close to the 58% incidence reported in an early study by Ling et al.14 and lower than the 71% incidence reported in the early study of Lewis and Smith.12
Among our 53 cases in which the corticosteroid was identified, prednisolone was administered in 20 (mean dose = 46.5 ± 28.5 mg/day, [n = 19]), prednisone in 16 (41.4 ± 24.7 mg/day [n = 15]), methylprednisone in seven (38.1 ± 44.3 mg/day), dexamethasone in seven (15.3 ± 6.2 mg/day [n = 6], with one outlier at 100 mg/day), betamethasone in two (2 mg/day and 4 mg/day), hydrocortisone in one (50 mg/day), and triamcinolone in one (80 mg/day + prednisone 10 mg/day). The mean (±SD) prednisone-equivalent dose, excluding the dexamethasone outlier, was 63.6 ± 46.2 mg/day, and the range 5–200 mg/day. This mean prednisone-equivalent dose supports the suggestion in the literature that psychiatric side-effects are more likely to occur at higher corticosteroid doses.
Focusing on individual syndromes and cases for which data are available, we found that delirium (n = 9) had a mean (±SD) onset of 7.0 ± 9.3 days after starting the corticosteroid (range 1–30 days), mean (±SD) prednisone-equivalent dose of 62.2 ± 45.5 mg/day, and mean (±SD) recovery time of 9.6 ± 9.3 days (range 1–25 days). For depression, mean onset was 12.4 ± 10.2 days (range 2–30 days) (n = 12), mean prednisone-equivalent dose 73.2 ± 48.4 mg/day (range 5–160 mg/day), and mean recovery time 20.9 ± 10.5 days (range 7–42 days) (n = 8 cases, with one still ill at time of reporting and three unknown). For hypomania/mania (n = 28), these statistics were: mean onset (n = 24), 14.0 ± 16.7 days (range 1–60 days), mean prednisone-equivalent dose (n = 25) 44.8 ± 45.6 mg/day, and recovery time (n = 19) 21.3 ± 18.8 days. Hypomania/mania seems to occur at somewhat lower doses than delirium. Compared to earlier reports, our findings indicate similar mean recovery times for delirium (several days), depression (4 weeks) and hypomania/mania (3 weeks). Without distinguishing various syndromes, Appenzeller et al.54 noted a median time to recovery of 13.3 ± 5.2 days in SLE patients suffering corticosteroid-induced psychoses. Some patients were treated with psychotropic drugs and some recovered simply by virtue of corticosteroid discontinuation.
The following case illustrates the difficulty that may be experienced in treating corticosteroid-induced depression, particularly when medical stressors continue and the corticosteroid cannot be quickly reduced. It also illustrates the need in such cases for prolonged use of psychotropic treatment, and the ultimate good prognosis. The authors received informed consent to publish her case.
Mrs S, an 85-year-old widowed, socially active woman with no prior psychiatric history, developed temporal arteritis with abrupt and permanent loss of vision in her right eye and blurred vision in her left. Subsequently, she began oral prednisone 60 mg/day. While tapering down to 40 mg/day 1 month later, she developed significant depressive and psychotic symptoms that resulted in her hospitalization for apparent steroid-induced psychosis. The psychosis resolved several weeks later and the patient was discharged, but continued to become increasingly depressed. Her depressive symptoms were marked anhedonia, apathy, and poor concentration, and she was disheveled in appearance with poor grooming and loss of function. Although her prednisone dose was lowered to 10 mg/day, she continued to decline. Four months later, after making suicidal statements and becoming assaultive toward her 24-h caregiver, she was hospitalized again, this time for almost 2 months. She was severely depressed, hopeless, nihilistic, with delusional guilt, visual hallucinations, poor insight and self-neglect. In hopes of saving the remaining vision in her left eye, methotrexate was added to her prednisone 10 mg/day. She was stabilized and discharged on olanzapine 7.5 mg alternating with 10 mg q.h.s., bupropion XL 300 mg/day, duloxetine 90 mg/day, benztropine 0.5 mg q.h.s., and, lorazepam 0.5 mg bid p.r.n. Two weeks later upon follow up, Mrs S was neatly groomed and cheerful. She denied depressed mood, anhedonia and suicidality. In the following months, her mood remained stable and she resumed social activities while continuing her hospital discharge medications. Olanzapine and duloxetine were gradually decreased, while bupropion, benztropine and lorazepam were discontinued after 4 months.
TREATMENT OF CORTICOSTEROID-INDUCED PSYCHIATRIC SYMPTOMS
The literature on the treatment of corticosteroid-induced psychiatric symptoms is limited to multiple case reports and a few small trials. These provide clinical guidance, but require larger double-blind, placebo-controlled trials to meet the standard of evidence-based medicine.
Many authors49,64,105 emphasize the importance of educating patients and their families about the risks of corticosteroid-induced psychiatric side-effects and of seeing patients soon after these drugs are begun, since these adverse effects may have rapid onset. The common occurrence of suicidal ideation (and less often, suicidal behaviors) must be kept in mind and preventive measures considered.106 When psychiatric symptoms occur, contributions of the underlying medical condition(s), other drugs or treatments, withdrawal from drugs such as alcohol and benzodiazepines, medical complications, such as infections, metabolic derangements or paraneoplastic syndromes, and contributions of psychosocial stressors, including the illness itself, will have to be taken into account and managed.
As has often been pointed out, treatment of corticosteroid-induced psychiatric symptoms should start whenever possible with dose reduction or stopping the drug.12,14,48,49,54,64,105 Without specifying the diagnoses associated with particular drugs, Lewis and Smith12 report that simply tapering the corticosteroid dose to zero resolved the psychiatric symptoms in 94% of 36 cases. Appenzeller et al. reported that simply tapering off the drug was effective in half their cases.54 Warrington and Bostwick105 cite endocrine experts107 who recommend tapering the corticosteroid dose to 40 mg/day of prednisone equivalent when discontinuation is not possible, followed as quickly as is safe by a taper to a physiological dose of 7.5 mg/day. The possibility of inducing psychiatric symptoms by tapering too quickly must be born in mind.14,72,75,108 Even slow taper, however, has led to the onset of depression in patients who were euthymic on prednisone, with the depression lasting 6–8 weeks after completing the taper.109
Corticosteroid-induced hypomania, mania and mixed mania have been successfully treated with a typical antipsychotic or mood stabilizer, most often haloperidol,18,53,54,76,102 haloperidol plus lithium,2 risperidone,3,91,93 quetiapine,95 olanzapine,24,92,110,111 olanzapine with valproate,18,112,113 carbamazepine,18 lithium,18,114,115 lamotrigine plus clonazepam,94 or clonazepam alone in a case where lithium had been ineffective.96 In some cases a combination of an antipsychotic and a benzodiazepine has been required. In a 5-week, open-label trial of olanzapine 2.5–20 mg/day (mean 8.5 mg/day) for mania or mixed mania symptoms secondary to corticosteroids, Brown et al.62 observed marked improvement in 11 of 12 outpatients; one patient withdrew for lack of efficacy. Renal function must be considered when contemplating the use of lithium and potential drug interactions are always a consideration.
Corticosteroid-induced depression has responded to lithium alone,77–79 lithium added to mianserin, amitriptyline,18 intravenous clomipramine,18 fluoxetine,80 venlafaxine,2 and, low-dose fluvoxamine.81 Psychotic depression has responded to electroconvulsive therapy2 and to combinations of sertraline, risperidone and lorazepam,9 or paroxetine, risperidone and lorazepam.2 In the 1970s, Hall et al.26,116 recommended against the use of tricyclic antidepressants after observing increased mood lability or symptoms consistent with delirium in some patients; others have also reported poor response.114,117 Selective serotonin reuptake inhibitors may be preferable due to their lower side-effect profile. Prednisolone, the active metabolite of prednisone, may increase plasma levels of fluvoxamine,81 so that smaller-than-usual antidepressant doses may be utilized.
Corticosteroid-induced delirium, like delirium of other causes, may respond to haloperidol4,100 or an atypical antipsychotic,99 although the addition of other agents, such as a benzodiazepine may be necessary.90,99
In one case,15 corticosteroid-induced panic disorder with agoraphobia responded within 2 weeks to tranylcypromine 20 mg/day coupled with behavioral therapy. In a second case, these symptoms responded within an unspecified interval to fluvoxamine and supportive psychotherapy.16
As noted earlier, corticosteroid-induced dementia resolves much more slowly following drug discontinuation than do other syndromes and may leave residual cognitive decrements. Although improvement may be apparent 1 month after discontinuation,28,45 deficits in learning and memory may persist for 6 months or more.27,30,45 Symptoms may also remit despite continued corticosteroid treatment. A patient who experienced a manic psychosis on 125 mg/day of cortisone had symptoms of impaired memory, attention and concentration 1 week after recovery while euthymic on 1 mg/day of dexamethasone, but normal mental status when examined 9 months later, despite being maintained in the interim on 0.5–1.5 mg/day of dexamethasone.45 Interestingly, memantine, used in treating Alzheimer's disease, has been reported in a double-blind, crossover trial of modest size to decrease adverse effects of corticosteroid treatment on declarative memory.118 A small, double-blind, placebo controlled study also suggested improvement in declarative memory as a result of treatment with lamotrigine, although the drug was not well-tolerated.119
For patients who have experienced a corticosteroid-induced psychosis or other severe adverse psychiatric effect, the clinician may wish to attempt a preventive intervention when the steroid is again needed. A number of strategies to prevent psychotic symptoms are described in case reports and one case series, but whether these symptoms would have returned absent these interventions is unknown. Falk et al.120 treated 27 patients receiving corticotrophin for multiple sclerosis or retrobulbar neuritis with prophylactic lithium carbonate (serum levels 0.8 to 1.2 mEq/l) and compared them to a historical control group not receiving concurrent lithium. Psychosis had occurred in 14% (6/44) of the historical control group, but was seen in none of the lithium-treated group. Two patients discontinued lithium for side-effects. Goggans et al.121 reported that a patient with a history of steroid psychosis had no recurrence during a second course of prednisone 60 mg/day preceded and accompanied by lithium 900 mg/day. In another case, lithium did not prevent the onset of mania with a first course of prednisone, but when prednisone was resumed 3 weeks later with concurrent lithium, the mania did not recur.122 A woman who had experienced two previous psychotic episodes followed by persistent melancholic depression safely underwent a third steroid course while treated with protriptyline, lithium and haloperidol (administered for only a few days).123 Bloch et al.83 reported that a patient who had become psychotic during two prior courses of corticosteroid treatment had no symptoms during a third course (methylprednisolone 1 g tapered to zero over 10 days, followed by prednisone 30 mg/day) so long as chlorpromazine 150 mg/day was co-administered. When it was tapered off, the patient became hypomanic and it had to be resumed. A patient who had experienced a manic episode when treated with prednisone 40 mg/day tolerated two subsequent courses while taking gabapentin 900 mg/day starting 1 day before resuming corticosteroids.89 In a bipolar patient whose lithium had to be discontinued because of worsening interstitial nephritis, lamotrigine plus clonazepam was effective in treating mania present on hospital admission and in preventing an exacerbation when the patient was placed on high-dose prednisone.94 Valproate has been reported to prevent steroid-induced psychosis in one case.112 In a double-blind, placebo-controlled trial of 30 patients, acetaminophen 4000 mg/day did not differ from placebo with respect to change in depressive symptoms.124
DIRECTIONS FOR FUTURE RESEARCH
Brown and Suppes57