REVIEW

Psychiatric Adverse Effects of Pediatric Corticosteroid Use Linda B. Drozdowicz, BS, and J. Michael Bostwick, MD Abstract Corticosteroids, highly effective drugs for myriad disease states, have considerable neuropsychiatric adverse effects that can manifest in cognitive disorders, behavioral changes, and frank psychiatric disease. Recent reviews have summarized these effects in adults, but a comprehensive review on corticosteroid effects in children has not been published since 2005. Here, we systematically review articles published since then that, we find, naturally divide into 3 main areas: (1) chronic effects of acute prenatal and neonatal exposure associated with prematurity and congenital conditions; (2) immediate behavioral effects of acute exposure via oncological protocols; and (3) acute behavioral effects of sporadic use in children and adolescents with other conditions. PsycInfo, MEDLINE, Embase, and Scopus were queried to identify articles reporting psychiatric adverse effects of corticosteroids in pediatric patients. Search terms included corticosteroids, adrenal cortex hormones, steroid psychosis, substance-induced psychoses, glucocorticoids, dexamethasone, hydrocortisone, prednisone, adverse effects, mood disorders, mental disorders, psychosis, psychotic, psychoses, side effect, chemically induced, emotions, affective symptoms, toxicity, behavior, behavioral symptoms, infant, child, adolescent, pediatric, paediatric, neonatal, children, teen, and teenager. Following guidelines for systematic reviews from the Potsdam Consultation on Meta-Analysis, we have found it difficult to draw specific conclusions that are more than general impressions owing to the quality of the available studies. We find a mixed picture with neonates exposed to dexamethasone, with some articles reporting eventual deficits in neuropsychiatric functioning and others reporting no effect. In pediatric patients with acute lymphoblastic leukemia, corticosteroid use appears to correlate with negative psychiatric and behavioral effects. In children treated with corticosteroids for noncancer conditions, adverse effects have been observed both during treatment and after cessation, although the data from article to article are not consistent enough to establish dose relationships. By and large, inhaled corticosteroids are considered safe and free of severe neuropsychiatric effects. Although both antipsychotic medications and benzodiazepines have been used to treat corticosteroid-induced mania and psychosis, no unified management strategy has emerged. Large-scale standardized investigations are needed to clarify the psychiatric effect of corticosteroids on children in all these conditions. Meanwhile, there is general agreement that patients as well as caregivers should be warned of the potential for behavioral adverse effects when patients receive these drugs. ª 2014 Mayo Foundation for Medical Education and Research

T

he prescription of corticosteroids is ubiquitous in many disease states, ranging from allergic, renal, and respiratory disorders to inflammatory, malignant, and dermatologic conditions. Like the metaphorical double-edged sword, these drugs offer symptomatic relief and halt disease progress while inducing troublesome adverse effects. Corticosteroids reduce inflammation by entering cells and binding as agonists to cytosolic glucocorticoid receptors. Once bound, these receptors transactivate the b-2 adrenergic receptor gene, the lipocortin-1 gene, the IL-10 gene, and the NF-kB inhibitor gene to produce anti-inflammatory effects.1 Many other genes

n

Mayo Clin Proc. 2014;89(6):817-834

are transactivated as well, and these are thought to produce the unwanted adverse effects. Over the 2-decade span of childhood, corticosteroids are first used neonatally in cases of prematurity, low birth weight, bronchopulmonary dysplasia (BPD), heart defects, and suspected congenital adrenal hyperplasia. Later, indications for systemic corticosteroids include malignant diseases such as acute lymphoblastic leukemia (ALL) and Hodgkin lymphoma (HL); autoimmune conditions such as Crohn disease, multiple sclerosis, and immune thrombocytopenic purpura; and oral surgical procedures (prevention of postoperative edema), Duchenne muscular dystrophy, nephrotic syndrome, and

Mayo Clin Proc. n June 2014;89(6):817-834 n http://dx.doi.org/10.1016/j.mayocp.2014.01.010 www.mayoclinicproceedings.org n ª 2014 Mayo Foundation for Medical Education and Research

From the Mayo Medical School (L.B.D.) and Department of Psychiatry, Mayo Clinic (J.M.B.), Rochester, MN.

817

MAYO CLINIC PROCEEDINGS

ARTICLE HIGHLIGHTS n

n

n

n

Whether neonatal exposure to dexamethasone is associated with chronic adverse psychiatric effects is unclear. Corticosteroid use in acute lymphoblastic leukemia correlates with negative psychiatric and behavioral effects. Acute corticosteroid exposure may produce adverse psychiatric events at any point during treatment and even after cessation. Dose-behavioral dyscontrol relationships cannot be calculated owing to the paucity of large comparable studies.

n

Inhaled corticosteroids appear generally safe and free of severe neuropsychiatric adverse effects.

n

Management strategies for psychiatric adverse reactions vary, but antipsychotic medications and/or benzodiazepines may be effective. Educating patients and parents will improve surveillance and decrease the likelihood of adverse events going unrecognized.

n

asthma. Among the extensive potential adverse effects of chronic use are osteoporosis, avascular necrosis of bones, Cushingoid body habitus, acne, hypertension, gastritis, amenorrhea, glaucoma, diabetes mellitus, and immunosuppression.2 These somatic adverse effects notwithstanding, this article focuses on the neuropsychiatric adverse effects that may accompany corticosteroid use. In the acute setting, these effects may include mood elevation ranging from mild euphoria to full-blown mania, insomnia, depression, and frank psychosis. With regard to psychosis, corticosteroid medications may mimic stress and alter the hypothalamicpituitary-adrenal axis, producing a state of heightened psychosis risk as described in the stress-vulnerability model of schizophrenia.3,4 The neuropsychiatric adverse effect profile of corticosteroids has been relatively well characterized and reviewed in adults, but the pediatric literature is less definitive.5 In adults, the most common adverse effects of short-term corticosteroid therapy are euphoria and hypomania, and symptoms tend to arise early in the treatment cycle. In contrast, long-term therapy tends to induce depressive symptoms. Dosage correlates with the incidence of adverse effects in adults, but dose is not related to the severity of adverse effects.5 818

Mayo Clin Proc.

n

As for pediatric data, in a 2005 article, the last to summarize the psychiatric adverse effects of corticosteroid use in children, Stuart et al6 stated the need for better studies to calculate incidence and prevalence. They suggested that it is difficult to establish a dose-response effect owing to the complex interplay of genetics, individual susceptibility, and environmental stressors.6 From the available evidence, they were unable to delineate clear risk factors, although some studies suggested that oral dexamethasone (DEX) may cause more adverse effects than do other corticosteroids.6 Of note, the article by Stuart et al considers cancer-associated corticosteroid use and noncancer applications as one category and it does not discuss neonatal exposure. This article expands the literature by reviewing the pediatric studies published since 2005, by delineating oncological protocols from other corticosteroid regimens, and by including articles covering prenatal and neonatal corticosteroid exposure. We focus on these areas because most of the articles written since 2005 fall into these categories. METHODS With the goal of performing a systematic review, an experienced research librarian conducted a literature search in June 2013 using PsycInfo, MEDLINE, Embase, and Scopus to identify English language journal articles reporting psychiatric adverse effects of corticosteroids in pediatric patients aged 0 to 18 years. The search terms corticosteroids, psychiatric effects, and pediatric populations were adjusted to each database to include database-appropriate subject headings, pharmacological actions, or thesaurus terms, and these terms were supplemented with keyword searches. Search terms included corticosteroids, adrenal cortex hormones, steroid psychosis, substance-induced psychoses, glucocorticoids, dexamethasone, hydrocortisone, prednisone, adverse effects, mood disorders, mental disorders, psychosis, psychotic, psychoses, side effect, chemically induced, emotions, affective symptoms, toxicity, behavior, behavioral symptoms, infant, child, adolescent, pediatric, paediatric, neonatal, children, teen, and teenager. Bibliographies from key articles were also searched, and Scopus was used to identify articles citing key articles. Owing to the relative scarcity of large-scale standardized studies, all types of research reports with any level of relevant data were

June 2014;89(6):817-834

n

http://dx.doi.org/10.1016/j.mayocp.2014.01.010 www.mayoclinicproceedings.org

PSYCHIATRIC EFFECTS OF PEDIATRIC CORTICOSTEROIDS

included in this review. As suggested in “Methodologic guidelines for systematic reviews of randomized control trials in health care from the Potsdam Consultation on Meta-Analysis .,” in the absence of highest-quality studies, the flawed studies that exist should nonetheless be reviewed for tentative findings.7 The lack of standardization in studies precluded the establishment of specific inclusion criteria. No restrictions were placed on article type or corticosteroid or formulation type. No specific definition of “neuropsychiatric adverse effects” was used to exclude articles before review. Articles written before 2005 were excluded because the last review article specifically focusing on children and adolescents was written in 2005.6 Of the 414 articles retrieved in the search, 369 were rejected either owing to date of publication (before 2005) or failure to discuss actual patient data (vs a completely theoretical framework). The remaining 45 articles were used for the review: 9 were randomized controlled trials (RCTs), 7 were prospective cohort studies, 11 were retrospective cohort studies, 3 were case series, 12 were case reports, 2 were cross-sectional studies, and 1 was a meta-analysis of RCTs.

for various conditions, finding a weightedaverage incidence of “severe” psychiatric reactions of 6%. Naber et al12 found that in a cohort of 50 patients treated with high-dose corticosteroids, 36% developed either mania or depression within days of therapy initiation. The discrepancies in calculated incidence could be attributable in part to poorly defined measures, such as the definition of “severe” in relation to psychiatric symptoms. In the pediatric studies we reviewed, few even attempted to calculate the incidence of psychiatric adverse effects. Many did not rise above the level of case reports. Attempting to combine such disparate reports to calculate the incidence of neuropsychiatric adverse effects in children is not appropriate. That said, we found 1 large study documenting incidence: Mitchell et al13 studied a group of more than 1500 children treated with corticosteroids during ALL therapy, finding a behavioral adverse effect incidence of 6% for DEX. These effects ranged from mood swings and lability to severe depression and violence toward self or others, with a trend for girls to develop depression and for boys to become aggressive.13

INCIDENCE Before discussing the incidence of psychiatric adverse events, it is important to consider the baseline prevalence of psychiatric illness. According to the National Institute of Mental Health, the prevalence of serious mental illness (excluding developmental and substance use disorders) in adults in 2011 was 4%.8 Data from the National Health and Nutrition Examination Survey indicate that in children aged 8 to 15 years, 13% had a diagnosable mental disorder in the past year.9 However, much of this was accounted for by attention deficit/hyperactivity disorder. Mood disorders affected 4% of this population, and anxiety disorders affected 1%.9 Behavioral issues such as temper tantrums are more common; 1 study examining 800 outpatient children aged 3 to 12 years found that 23% had temper tantrums, with 96% of these events occurring in 3- to 8-year-olds.10 The literature on corticosteroids exhibits considerable variability in reported incidences for psychiatric adverse effects in adults. In one study, Lewis and Smith11 performed a metaanalysis of uncontrolled studies involving a total of 2555 adult patients treated with corticosteroids

ADVERSE EVENTS

Mayo Clin Proc. n June 2014;89(6):817-834 www.mayoclinicproceedings.org

n

Neonatology Prematurity and diseases of the newborn have always caused significant morbidity. However, the use of corticosteroids in neonatology has curbed some of that vulnerability, particularly from pulmonary complications. These drugs have indications not only for BPD but also for congenital adrenal hyperplasia, heart defects, and other conditions. Understandably, physicians and families worry about the ultimate neurodevelopmental outcomes of newborns treated with corticosteroids. Bronchopulmonary dysplasia, also known as neonatal chronic lung disease, increases both morbidity and mortality in newborns.14 Mediated in part by persistent inflammation of the lungs stemming from prematurity, mechanical ventilation, and infection, among other causes, BPD may be attenuated or prevented with corticosteroids, particularly DEX.15 Some studies link acute prenatal and neonatal DEX exposure with chronic neurobehavioral sequelae, such as cerebral palsy, impaired social skills, or impaired motor skills (Table 1).16-19 One study

http://dx.doi.org/10.1016/j.mayocp.2014.01.010

819

820

TABLE 1. Summary of Neonatal Studies. Reference, year

Cohort characteristic/exclusion criteria

Cohort size/age at follow-up

Design

Relevant conclusions

Neonatal studies reporting neurodevelopmental or behavioral effects of corticosteroids Hirvikoski et al,18 2007

d

Children born between 1985 and 1995 who were prenatally treated with DEX owing to risk for congenital adrenal hyperplasia

d

d

61: 26 Children treated with DEX; 35 age- and sexmatched controls Assessed at the age of 7-17 y

d d

d

Retrospective study Fetuses at risk for CAH were treated with DEX from postmenstrual week 6 to week 7 to prevent virilization 20 mg/kg per day (up to 1.5 mg) depending on the pre-pregnancy weight

d

d

d

Karemaker et al,25 2006

d

d

Mayo Clin Proc. n

June 2014;89(6):817-834

Kutschera et al,16 2005

d

d

n

d

d

d

d

192: 46 Treated with DEX; 52 treated with hydrocortisone; 51 treated with betamethasone; 43 controls Assessed at 7-10 y

d

66: 33 Children treated with DEX; 33 controls Assessed at the age of 3-7 y

d

d

d

Retrospectively matched cohort study DEX began at 0.5 mg/kg per day and tapered off to 0.1 mg/kg per day over 21 d. Hydrocortisone began at 5 mg/kg per day and tapered off to 1 mg/kg per day over 22 d. Antenatal betamethasone (12 mg) administered to the mother twice intramuscularly.

d

Patients matched retrospectively with the control group DEX began with a total dose of 2.35 mg/kg between 7 and 28 d of life and lasted 7 d

d

Teacher’s Report Form displayed a significantly higher percentage of children in the DEX group with internalizing (withdrawn) behavioral problems according to clinical criteria than in the other groups (P

Psychiatric adverse effects of pediatric corticosteroid use.

Corticosteroids, highly effective drugs for myriad disease states, have considerable neuropsychiatric adverse effects that can manifest in cognitive d...
687KB Sizes 0 Downloads 4 Views