1. Introduction
The diagnosis of bipolar disorder (BD) in children and adolescents has been a controversial topic [Reference Goldstein, Birmaher, Carlson, DelBello, Findling and Fristad1], with much concern about risks of treating it [Reference Singh, Ketter and Chang2], with concerns about the harms of antipsychotic agents in particular [Reference Schneider, Taylor, Zalsman, Frangou and Kyriakopoulos3]. An alternative to antipsychotic agents would be mood-stabilizing drugs like lithium, yet clinicians also are reluctant to use that agent, especially with apprehension regarding cognitive side effects [Reference Findling4], as well as about long-term medical risks, such as hypothyroidism and chronic renal insufficiency [Reference McKnight, Adida, Budge, Stockton, Goodwin and Geddes5]. Further, many clinicians seem to be sceptical about the efficacy of lithium in children.
This paper seeks to shed light on these concerns, with the first systematic review on the safety and efficacy of lithium in children and adolescents with BD.
2. Materials and methods
As done before [Reference Amerio, Galvéz, Odone, Dalley and Ghaemi6, Reference Marchesi, Ossola, Amerio, Daniel, Tonna and De Panfilis7], this review was conducted according to methods recommended by the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [Reference Higgins and Green8, Reference Liberati, Altman, Tetzlaff, Mulrow, Gøtzsche and Ioannidis9].
2.1. Information sources and search strategy
Studies were identified searching the electronic databases MEDLINE, Embase, PsycINFO and the Cochrane Library. We combined the search strategy of free text terms and exploded MESH headings for the topic of treatment with lithium in children and adolescents combined as following: ((((((Lithium) OR Lithium carbonate) OR Lithium carbonate[MeSH Terms])) AND (((Children) OR Adolescent) OR Adolescent[MeSH Terms])) AND (((((((Bipolar disorder) OR BD) OR Bipolar) OR Manic depressive disorder) OR Manic depressive) OR Manic) OR Bipolar disorder[MeSH Terms])) AND ((((treatment*) OR therap*) OR pharmacotherap*) OR Therapeutics[MeSH Terms]). The strategy was first developed in MEDLINE and then adapted for use in the other databases (Appendix A in Supplementary material). Studies published in English through June 30th 2018 were included. In addition, further studies were retrieved from reference listing of relevant articles and consultation with experts in the field.
2.2. Inclusion criteria
2.2.1. Study population and study design
We considered studies that included children and adolescents with BD treated with lithium both in monotherapy and in combination with others psychotropic drugs. BD was considered if diagnostic criteria used were specified. Studies conducted on youths with different disorders than BD (e.g. dysphoric mood dysregulation disorder) were excluded (i.e. [Reference Dickstein, Towbin, Van Der Veen, Rich, Brotman and Knopf10–Reference Malone, Delaney, Luebbert, Cater and Campbell12]:). Participants of both sexes younger than 18 years of age were considered. Studies conducted on subjects with physical comorbidities such as epilepsy were excluded as non-representative of the study population [Reference Erwin, Gerber, Morrison and James13].
Among hospital-based studies, inpatients, day-hospital and outpatient subjects were included, while emergency care records were excluded as non-representative. All experimental and observational study designs were included apart from case reports and case series. Narrative and systematic reviews, letters to the editor, and book chapters were excluded.
2.2.2. Outcomes
The primary outcome was lithium effectiveness in children and adolescents with BD. Secondary outcomes were i) starting dose and dosing strategy, ii) brain-to-serum lithium association, and iii) safety and tolerability of lithium.
2.2.3. Study selection and data extraction
Identified studies were independently reviewed for eligibility by two authors (AA, FS) in a two-step process: A first screening was performed based on title and abstract, and then full texts were retrieved for a second screening. At both stages disagreements by reviewers were resolved by consensus. Data were extracted by two authors (AA, FS) and supervised by a third author (SNG) using an ad-hoc developed data extraction spreadsheet. The data extraction spreadsheet was piloted on 10 randomly selected papers and modified accordingly.
3. Results
Two hundred and twelve potential studies were identified from the selected databases and after cross-checking references of relevant articles. After removing duplicates, 152 articles were retrieved. Studies were screened and selected on the basis of pre-specified inclusion and exclusion criteria (Fig. 1). The search identified 30 articles that were included in the systematic review.
Flow diagram of selected articles.
*Search strategy limited to June 2018, English language, human subjects younger than 18 years old, and clinical trial.
3.1. Included studies
The characteristics of included studies are reported in Table 1. Twelve (40%) of the 30 studies were randomized controlled trials (RCTs) of which only one was longer than 6 months in duration. Most studies (n = 19, 63%) were short-term (8 weeks or less), while 4 studies (13%) provided long-term data of 6 months or longer. The smallest study included 6 subjects while the largest considered a sample of 279 subjects. The majority of the studies were conducted in North America (N = 28, 93%). In all the considered studies, diagnosis were based on the Diagnostic and Statistical Manual (DSM) criteria and were established using validated assessment scales (Table 1).
BD: Bipolar disorder; PMDD: Prepubertal major depressive disorder; MDD: Major Depressive Disorder; DBD: Disruptive behavior disorder; SDD: Substance dependency disorders; EUCD: Emotionally Unstable Character Disorder; ADD: Attention Defict Disorder; CD: Conduct disorder; Li+: lithium carbonate; DVPX: divalproex sodium; CBZ: carbamazepine; MPH: Methylphenidate; SGA: Second generation antipsychotic; DSM: Diagnostic and Statistical Manual of Mental Disorders; TEAM: Treatment of Early Age Mania; K-SADS: Kiddie Schedule for Affective Disorders and Schizophrenia; K-SADS-P: Kiddie Schedule for Affective Disorders and Schizophrenia for Psychopatology; K-SADS-E: Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children-Epidemiologic Version; K-SADS-PL: Schedule for Affective Disorders and Schizophrenia for School Age Children-Present and Lifetime Version; WASH-U-KSADS: Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia; CGI: Clinical Global Impression Scale; CGI-I: Clinical Global Impression-Improvement Scale; CGI-S: Clinical Global Impression-Severity Scale; CGI-BD: Clinical Global Impressions Improvement Scale for Bipolar Disorder; CGI-BP-IM: Clinical Global Impressions for Bipolar Disorder Improvement-Mania; YMRS: Young Mania Rating Scale; CDRS-R: Children’s Depression Rating Scale Revised; CGAS: Children’s Global Assessment Scale; BPRS: Brief Psychiatric Rating Scale; HAM-D: Hamilton Depression Rating Scale; ABC-C: Aberrant Behavior Checklist – Community Edition; VAS: Visual analog scale for behavior; VABS: Vineland Adaptive Behavior Scale; FH-RDC: Family History Research Diagnostic Criteria; SMD: Severe Mood Dysregulation; PANSS: Positive and Negative Syndrome Scale; CPT: Continuous Performance Test; K-PAL: Kinsbourne’s computerized version of the Paired Associated Learning paradigm; ACTeRS: Attention Deficit Disorder-Hyperactivity Comprehensive Teacher Rating Scale; IGRS: Inpatient Global Rating Scale; OAS: Overt Aggression Scale; MAOS: Modified Overt Aggression Scale; CBRS: Conners Bahavior Rating Scale; CTRS: Conners Teacher Rating Scale; RT: Reaction Time Task; CSRS: Conner's Symptom Rating Scale; MFF: Matching Familiar Figures: STRM: Short-term Recognition Memory; CAT: Concept Attainment Task; PANESS: Physical and Neurological Examination for Soft Signs; MPA: Minor Physical Anomalies; WISC-R: Wechsler Intelligence Scale for Children Revised; CPRS: Children’s Psychiatric Rating Scale; CTQ: Conners Teacher Questionnaire; PTQ: Parent Teacher Questionnaire; TORSA: Timed Objective Rating Scale for Aggression; DOTES: Dosage Record and Treatment Emergent Symptoms; GCCR: Global Clinical Consensus Rating; TESS: Treatment Emergent Symptoms Scale; GCJCS: Global Clinical Judgements Consesus Scale; POMS: Profile of Mood States; TSCRS: Teacher’s Self Control Rating Scale; Mini-Kid: Mini International Neuropsychiatric Interview for Children and Adolescents; CGI-BP-IM: Clinical Global Impressions for Bipolar Illness-Improvement Mania Scale; DICA-R: Diagnostic Interview for Children and Adolescents Revised; GAS: Global Assessment Scale (for subjects who were over 16 years of age at follow-up); MSEFCA: Modified Side Effect Form for Children and Adolescent; ALSES: Acute Lithium Side Effects Scale; FH-RDC: Family History Research Diagnostic Criteria; CARS: Children’s Affective Rating Scale; AIMS: Modified Abnormal Involuntary Movement Scale; PSSAC-R: Psychosocial Schedule for School-Age Children Revised; WASI: Wechsler Abbreviated Scales of Intelligence; CDRS-R: Children’s Depression Rating Scale-Revised; RCT: Randomized controlled trial.
3.2. Outcomes
Selected studies included children and adolescents with BD treated with lithium. Both lithium monotherapy and lithium in combination with adjunctive agents were included. Data about starting dose and dosing strategy, brain-to-serum lithium association, safety and tolerability were also reported (Table 2).
BD: bipolar disorder; EUCD: emotionally unstable character disorder; ADD: attention defict disorder; CD: conduct disorder; ADHD: attention deficit/hyperactivity disorder; SMD: Severe mood dysregulation; SDD: substance dependency disorders; Li+: lithium carbonate; DVPX: divalproex sodium; CBZ: carbamazepine; BDZ: benzodiazepines; AP: Antipsychotics; AD: Antidepressants; MS: Mood stabilizers; MPH: Methylphenidate; NAA: N-acetyl-aspartate; YMRS: Young Mania Rating Scale; CPRS: Children’s Psychiatric Rating Scale; GCJCS: Global Clinical Judgments Consensus Scale; BRMS: Bech–Rafaelsen Mania Scale; ABC-C: Aberrant Behavior Checklist – Community Edition; CGI: Clinical Global Impression Scale; OAS: Overt Aggression Scale; VABS: Vineland Adaptive Behavior Scale; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; CPT: Continuus Performance Test; K-PAL: Kinsbourne’s computerized version of the Paired Associated Learning paradigm; PSSAC-R: Psychosocial Schedule for School-Age Children Revised; CGAS: Children’s Global Assessment Scale; CDRS-R: Children’s Depression Rating Scale Revised; MTD: maximum tolerated dosage; GI: Gastrointestinal; TSH: thyroid stimulating hormone; SD: standard deviation; RR: response ratio; pts: patients; RT: reaction time; NS: Not specified; Vs.: Versus; EP: evoked potentials; NNT: Number needed to treat; *Related to the use of lithium; RCT: Randomized controlled trial; Differences statistically significant (p < 0.05).
4. Bipolar illness
Thirty studies assessed the use of lithium in children and adolescents with BD (Table 2). The majority of the selected studies (N = 22/30, 73%) were conducted on BD patients treated with lithium monotherapy. Eleven studies (37%) were specific for BD-I patients.
4.1. Lithium monotherapy
4.1.1. Manic or mixed episodes
Three RCTs reported improvements in manic or mixed symptoms and overall functioning in manic BD children and adolescents with lithium treatment [Reference Findling, Kafantaris, Pavuluri, McNamara, McClellan and Frazier14, Reference Findling, Robb, McNamara, Stansbrey and Calabrese20, Reference Landersdorfer, Findling, Frazier, Kafantaris and Kirkpatrick21]. Over 50% of patients met response and remission criteria in one out of three of the cited studies [Reference Findling, Kafantaris, Pavuluri, McNamara, McClellan and Frazier14]. These results were supported by three prospective non-randomized cohort studies [Reference Kafantaris, Dicker, Coletti and Kane18, Reference Kafantaris, Coletti, Dicker, Padula and Kane22, Reference Strober, Morrell, Burroughs, Lampert, Danforth and Freeman23] and one retrospective cohort study [Reference Kafantaris, Coletti, Dicker, Padula and Pollack24]. No significant difference in exacerbation rates between subjects treated with lithium and those switched to placebo was detected in only one RCT with a very short stabilization period (two weeks) [Reference Kafantaris, Coletti, Dicker, Padula, Pleak and Alvir25].
Considering bipolar subgroups, lithium effectiveness for manic symptoms was greater in adolescent-onset compared to prepubertal-onset patients in one study [Reference Strober, Morrell, Burroughs, Lampert, Danforth and Freeman23]. Manic adolescents with comorbid attention deficit hyperactivity disorder (ADHD) showed less robust and slower improvement with lithium compared to non-comorbid patients, both in a randomized and in a non-randomized trial [Reference Vitiello, Riddle, Yenokyan, Axelson, Wagner and Joshi26, Reference Strober, DeAntonio, Schmidt-Lackner, Freeman, Lampert and Diamond27]. In patients with substance abuse and BD, lithium was an effective for both conditions in one RCT [Reference Geller, Cooper, Sun, Zimerman, Frazier and Williams15].
4.1.2. Depressive episodes
A 6-week prospective non-randomized cohort study in BD-I depressed children and adolescents treated with lithium reported response and remission rates of 48% and 30%, respectively, with a large reduction in Children’s Depression Rating Scale-Revised (CDRS-R) scores (standardized effect size Cohen’s d = 1.7) [Reference Patel, DelBello, Bryan, Adler, Kowatch and Stanford28].
4.1.3. Prophylaxis
Three prospective non-randomized cohort studies reported long-term positive response to lithium treatment [Reference DeLong and Aldershof29, Reference Strober, Morrell, Lampert and Burroughs30], especially in those who responded to acute treatment with lithium [Reference Findling, Kafantaris, Pavuluri, McNamara, Frazier and Sikich31].
In a 18-month prospective non-randomized cohort study, 35% (N = 13/37) of patients who discontinued prophylactic lithium therapy showed nearly three times higher relapse rates compared to patients who continued lithium prophylaxis [Reference Strober, Morrell, Lampert and Burroughs30]. Early relapse was associated with a greater risk of future relapse.
In contrast, an Indian prospective non-randomized cohort study found that 64% (N = 16/25) of lithium-treated subjects relapsed after 18 ± 16.4 months (mean total follow-up duration 51.6 ± 4.1 months). The majority of the relapses (72.4%) occurred during prophylactic treatment. 28% (N = 7/25) and 36% (N = 9/25) of relapsing patients had single and multiple relapses respectively, with manic episodes being the most common polarity (N = 14/25, 58%) [Reference Jairam, Srinath, Girimaji and Seshadri32].
4.1.4. Offspring of manic-depressive patients
In a small RCT, 33% (N = 2/6) children, who met DSM-III criteria for BD and were offspring of manic-depressive patients, responded to lithium on both child and parent ratings and showed augmentation of evoked potentials (EPs) similar to what is seen in adults treated with lithium [Reference McKnew, Cytryn, Buchsbaum, Hamovit, Lamour and Rapoport33].
4.1.5. Lithium monotherapy vs. other psychotropic drugs
Five RCTs compared lithium monotherapy to other mood stabilizers or antipsychotics [Reference Findling, McNamara, Youngstrom, Stansbrey, Gracious and Reed34–Reference Walkup, Wagner, Miller, Yenokyan, Luby and Joshi38]. In a 18-month RCT, divalproex was not found to be superior to lithium as maintenance treatment in BD youths who had been stabilized on combined lithium plus divalproex for four weeks [Reference Findling, McNamara, Youngstrom, Stansbrey, Gracious and Reed34].
In a prospective non-randomized cohort study, lithium, divalproex, and carbamazepine all showed a large and similar effect size in treatment of acute manic or mixed episodes [Reference Kowatch, Suppes, Carmody, Bucci, Hume and Kromelis36].
Compared to antipsychotics for initial treatment of acute manic or mixed episode in children and adolescents, risperidone was more efficacious than mood stabilizers but had more metabolic side effects [Reference Geller, Luby, Joshi, Wagner, Emslie and Walkup35, Reference Salpekar, Joshi, Axelson, Reinblatt, Yenokyan and Sanyal37]. Also, risperidone was more effective than lithium or divalproex for children with BD-I who were nonresponders or partial responders to another prior antimanic agent [Reference Walkup, Wagner, Miller, Yenokyan, Luby and Joshi38].
4.2. Lithium in combination with adjunctive agents
As in adults, children and adolescents with BD frequently required long-term combination therapy [Reference Kowatch, Sethuraman, Hume, Kromelis and Weinberg39]. Therefore, adherence became important [Reference Drotar, Greenley and Demeter19]. A 6-month prospective non-randomized cohort study demonstrated that lithium or divalproex plus risperidone were equally efficacious and safe for manic and mixed symptoms in pediatric mania [Reference Pavuluri, Henry, Carbray, Sampson, Naylor and Janicak40]. In psychotic mania, one study found that adjunctive antipsychotic medication needed to be maintained longer than 4 weeks in the majority of adolescents [Reference Kafantaris, Coletti, Dicker, Padula and Kane17].
As demonstrated by two prospective non-randomized cohort studies, the combination of lithium and divalproex was effective in treating acute manic and depressive symptoms in juvenile BD and in restabilizing BD patients who had treated with combined lithium plus divalproex sodium but later relapsed with monotherapy of one agent [Reference Findling, McNamara, Gracious, Youngstrom, Stansbrey and Reed41, Reference Findling, McNamara, Stansbrey, Gracious, Whipkey and Demeter42].
5. Starting dose and dosing strategy
In patients with BD-I, two phases in the distribution of lithium was observed, with an initial half-life of 2.4 h and a later half-life of 27 h [Reference Findling, Landersdorfer, Kafantaris, Pavuluri, McNamara and McClellan43]. Multiple dose simulations suggested that a starting dose of 300 mg once daily for those weighing less than 30 kg, and 300 mg twice or three times daily for youths weighing 30 kg or more, appear to be appropriate based on safety margins for trough concentrations. Later trials from the same research group observed a good tolerability of starting dose of 900 mg/day for most subjects [Reference Findling, Kafantaris, Pavuluri, McNamara, McClellan and Frazier14, Reference Findling, Robb, McNamara, Stansbrey and Calabrese20].
6. Brain-to-serum lithium association
A cross-sectional study conducted on BD-I patients showed a positive correlation between serum and brain lithium concentrations, with younger subjects having lower brain-to-serum concentration ratios than adults (0.58, SD = 0.24 vs. 0.92, SD = 0.36) [Reference Moore, Demopulos, Henry, Steingard, Zamvil and Katic44]. According to these results, children and adolescents may need higher maintenance serum lithium concentrations than adults to reach similar brain lithium concentrations.
7. Safety and tolerability
Lithium’s recommended target dose in children and adolescents is 30 mg/kg/day, with 0.6–1.2 mEq/L serum levels [Reference Martin, Charney and Leckman45] even though other authors [Reference Findling, Kafantaris, Pavuluri, McNamara, McClellan and Frazier14, Reference Findling, Robb, McNamara, Stansbrey and Calabrese20] suggest more aggressive dosing, particularly for acute manic or mixed episodes. For instance, the FDA recommends dosing lithium until therapeutic response or a maximal blood level of 1.4 mEq/L is achieved or a dose-limiting side effect is present. Studies in this review tended to be within those guidelines, with 10–30 mg/kg/day dosages range, 0.6–1.5 mEq/L serum levels (Table 1). Most of the selected studies reported side effects in the mild to moderate range, with low dropout rates. No serious adverse events were reported either with lithium monotherapy or in combination with other psychotropic drugs (Table 1). Obviously, close serum level monitoring is required to ensure that lithium is safe and well-tolerated.
The most common side effects of lithium in children and adolescents were gastrointestinal symptoms (including nausea, vomiting, diarrhea, abdominal cramps, stomach pain), tremor, polyuria, polydipsia, and enuresis (Table 1). Only a minority of patients presented hypothyroidism with increased thyroid stimulating hormone [Reference Findling, Robb, McNamara, Stansbrey and Calabrese20] and/or thyroid enlargement [Reference Findling, Kafantaris, Pavuluri, McNamara, McClellan and Frazier14, Reference DeLong and Aldershof29]. Most of the studies report an increase in appetite and weight gain, which is a problem with several other BD agents [Reference Findling and Chang46]. However, this difference was not statistically different from placebo [Reference Findling, Robb, McNamara, Stansbrey and Calabrese20]. No cases of acute kidney injury or chronic kidney disease were reported.
8. Discussion
This study is the first systematic review to look specifically at the use of lithium in children and adolescents with BD. Thirty studies were included. Almost sixty percent of subjects were studied in RCTs (n = 12 studies) that evaluated lithium efficacy in acute manic-mixed episodes, and as a prophylactic agent. Lithium was effective in monotherapy and in combination with antipsychotics for treating acute manic and mixed episodes. Lithium showed a greater response in manic episodes without psychotic symptoms, and in the absence of ADHD comorbidity. Adolescents had a better response than prepubescent children, and prophylactic efficacy was superior to divalproex and proportional to acute response [Reference Findling, Kafantaris, Pavuluri, McNamara, Frazier and Sikich31]. Lithium also was effective in treating BD with secondary substance use disorder.
It is unclear whether lithium is effective in treating bipolar depression in children and adolescents. The only study that evaluated lithium efficacy on bipolar depression in youth [Reference Patel, DelBello, Bryan, Adler, Kowatch and Stanford28] found relatively low response rates that are in line with previous results in depressed children with a family history of BD [Reference Geller, Cooper, Zimerman, Frazier, Williams and Heath16] and in adults [Reference Selle, Schalkwijk, Vazquez and Baldessarini47].
Lithium appears generally safe for use by children and adolescents with only mild-moderate side effects. Its long-term risks also do not appear to be notable, at least as identified so far, and seem consistent with the adult literature. No cases of renal failure were observed in the sample of over 2000 subjects included in this systematic review, but this is probably explained by the short follow-up. In fact, accordingly to the literature, long-term patients (generally after 10–20 years of treatment) have an increased risk to develop impaired renal function due to a slowly progressive chronic interstial nephritis. Hypothyroidism was observed, though only in a minority of patients.
All the RCTs that found a lower efficacy of lithium compared to risperidone [Reference Vitiello, Riddle, Yenokyan, Axelson, Wagner and Joshi26, Reference Geller, Luby, Joshi, Wagner, Emslie and Walkup35, Reference Salpekar, Joshi, Axelson, Reinblatt, Yenokyan and Sanyal37, Reference Walkup, Wagner, Miller, Yenokyan, Luby and Joshi38] were from the Treatment of Early Age Mania Study (TEAM).
8.1. Epidemiological and clinical background of BD in children and adolescents
In the past, with the exception of a few experts, like Kraepelin and Ziehen, bipolar illness in children essentially was not considered a diagnostic possibility until recently. Estimated prevalence in adolescence varies from 1.8% [Reference Van Meter, Moreira and Youngstrom48] to 2.5% [Reference Merikangas, Cui, Kattan, Carlson, Youngstrom and Angst49], being slightly higher in the US than in Europe. In 2011, the World Health Organization listed bipolar spectrum disorder as the fourth leading cause of disability among adolescent ages 15–19 years worldwide [Reference Gore, Bloem, Patton, Ferguson, Joseph and Coffey50]. Annual rates of BD diagnosis [Reference Blader and Carlson51] and related hospitalization [Reference Shapiro, Timmins, Swampillai, Scavone, Collinger and Boulos52] in youth are increasing. Compared to adult-onset BD, children and adolescents experience more symptoms, comorbidities and mood switches and have a poorer prognosis [Reference Baldessarini, Tondo, Vazquez, Undurraga, Bolzani and Yildiz53, Reference Holtzman, Miller, Hooshmand, Wang, Chang and Hill54]. Because BD is a recurrent condition with more than 70% of subjects relapsing by early adulthood [Reference Geller, Tillman, Bolhofner and Zimerman55], it is important to treat prophylactically even in childhood, when it is often misdiagnosed [Reference Baethge, Glovinsky and Baldessarini56] or left untreated [Reference Khazanov, Cui, Merikangas and Angst57].
8.2. Comparison with FDA recommendations and main guidelines
In the paediatric population, lithium is the only mood stabilizer with a US Food and Drug Administration (FDA) indication for acute mania and maintenance treatment of BD and it is licensed for treatment of acute mania in the UK [Reference Giles and Martini58, Reference Goodwin, Haddad, Ferrier, Aronson, Barnes and Cipriani59]. Clinicians still display qualms about lithium [Reference Ko, Swampillai, Timmins, Scavone, Collinger and Goldstein60, Reference Strejilevich, Urtueta-Baamonde, Teitelbaum, Martino, Marengo and Igoa61], with BD children and adolescents receiving complex treatment regimens, often involving multiple psychotropic drugs; only 2% receive lithium monotherapy [Reference Dusetzina, Weinberger, Gaynes, Farley, Sleath and Hansen62].
Effective treatments in BD youths are still withheld or underused. More than one third of children with BD receive two or more medications belonging to different classes (i.e. mood stabilizers, antipsychotics, antidepressants or stimulants) feeding the problem of off-label prescriptions [Reference Sharma, Arango, Coghill, Gringras, Nutt and Pratt63].
The US FDA also has approved risperidone, aripiprazole, quetiapine, olanzapine, olanzapine in association with fluoxetine, lurasidone and asenapine for the treatment of BD in youth; the higher efficacy of risperidone above lithium in the two trials available might have been driven by the high comorbidity rates in the study populations, or it may reflect more rapid acute onset of effect [Reference Giles and Martini58]. Aripiprazole, the only antipsychotic licensed in the UK for paediatric BD, is comparable in efficacy to traditional mood stabilizers [Reference Meduri, Gregoraci, Baglivo, Balestrieri, Isola and Brambilla64].
8.3. Tolerability
Many second-generation antipsychotics are associated with a greater burden of metabolic side effects that, along with the recent warning about the use of antipsychotic in children and adolescents, could argue in favour of mood stabilizers [Reference Correll, Detraux, De Lepeleire and De Hert65–Reference Samaras, Correll, Mitchell and De Hert67]. In fact, lithium side effects (e.g. nausea, hand tremor, thirst and polyuria, diarrhoea) are dose-dependent and, although relatively frequent [Reference Licht, Vestergaard, Kessing, Larsen and Thomsen68], are only mild-moderate in effect size. Cognitive effects also need to be considered, with a recent study finding impairment in executive control in adolescents treated with lithium [Reference Lera-Miguel, Andres-Perpina, Fatjo-Vilas, Fañanás and Lázaro69]; another prior study did not find such cognitive impairment though [Reference Muralidharan, Kozicky, Bucker, Silveira, Torres and Yatham70]. Lithium treatment in adolescents is associated with an increase in blood TSH levels in up to 25% of the subjects [Reference Amitai, Zivony, Kronenberg, Nagar, Saar and Sever71]. In this systematic review, only one subject had clinically significant thyroid impairment.
It also is important to note that no cases of acute kidney injury or chronic kidney disease were observed. Still, serial monitoring of renal and thyroid functioning is recommended [Reference Amitai, Zivony, Kronenberg, Nagar, Saar and Sever71, Reference Scahill, Farkas and Hamrin72].
In sum, special concern about the use of lithium in children and adolescents, at least on safety grounds, seems unwarranted based on this systematic review.
9. Limitations
The main limitation of this systematic review is linked to the characteristics of the selected studies, such as brevity of follow-up periods, small sample sizes, and analysis strategies (Table 1). Small sample sizes and enrolment of subjects mainly from US sites (almost 94%) may limit generalizability. Potential confounding factors in these studies include demographic and historical illness variables in non-randomized trials, which often were not appropriately analysed through multivariate modelling. The inclusion of only BD studies might limit the results regarding lithium tolerability in children and adolescents [Reference Campbell, Silva, Kafantaris, Locascio, Gonzalez and Lee11, Reference Malone, Delaney, Luebbert, Cater and Campbell12].
The main strength of this systematic review is linked to the inclusion of 12 RCTs (Table 1), which would limit the effect of confounding variables Further, this review was systematic, including the entire scientific evidence published so far. Also, diagnoses were consistently based on DSM criteria and were established by trained investigators using validated assessment scales mainly with interrater reliability. This is particularly relevant because one of the main features of BD is mood instability and it is easy to attribute mood swings to the behavioural vicissitudes of adolescence or to psychosocial changes [Reference Harrison, Cipriani, Harmer, Nobre, Saunders and Goodwin73]. Moreover the comorbidity rate in BD in youth is more than 80% [Reference Birmaher, Axelson, Goldstein, Strober, Gill and Hunt74–Reference Kowatch, Youngstrom, Danielyan and Findling76], with up to 50% for anxiety disorders [Reference Frias, Palma and Farriols77]. Thus, in mild cases, adjustment disorders are frequently diagnosed [Reference Kessing, Vradi, McIntyre and Andersen78]. The debate about the possibility of broadening the diagnosis of BD in youth to a larger spectrum remains open [Reference Birmaher, Gill, Axelson, Goldstein, Goldstein and Yu79]. Future studies should better define this diagnostic spectrum in order to allow better study of drug efficacy.
10. Clinical implications
The results of this systematic review, showing efficacy and safety of lithium in children with BD, are in line with clinical recommendations in literature. According to guidelines for children and adolescents with BD [Reference McClellan, Kowatch and Findling80, Reference Kowatch, Fristad, Birmaher, Wagner, Findling and Hellander75, Reference Kowatch, Youngstrom, Danielyan and Findling76], lithium is considered a first line treatment and can be prescribed in monotherapy for up to 8 weeks, if there are no psychotic features. If there are psychotic features, treatment combining a mood stabilizer and an antipsychotic is recommended.
Lithium should be considered if there is a comorbid substance use disorder, which has high frequency in this population and portends poor prognosis [Reference Dickstein, Towbin, Van Der Veen, Rich, Brotman and Knopf10].
Overall, these results in children somehow overlap with those in adults which find efficacy with lithium [Reference Miura, Noma, Furukawa, Mitsuyasu, Tanaka and Stockton81, Reference Lähteenvuo, Tanskanen, Taipale, Hoti, Vattulainen and Vieta82], benefit with lithium in preventing suicide [Reference Cipriani, Hawton, Stockton and Geddes83], and general tolerability of lithium [Reference McKnight, Adida, Budge, Stockton, Goodwin and Geddes5]. Similarly, when accounting for body size, the pharmacokinetic parameters in paediatric patients were within the range of estimates from adults [Reference Landersdorfer, Findling, Frazier, Kafantaris and Kirkpatrick21]. As in adults [Reference Biel, Peselow, Mulcare, Case and Fieve84], lithium discontinuation in BD after successful maintenance monotherapy is not advisable.
Moreover, children who are treated with lithium are less likely to show mood instability, impulsivity and self-injurious behaviour, identity confusion, and interpersonal problems [Reference Ko, Swampillai, Timmins, Scavone, Collinger and Goldstein60], all of which are poor prognostic factors [Reference Miklowitz85].
To date, the field of child and adolescent psychiatry lacks validated prophylactic therapy for depressive and bipolar illness, and little is known about the benefit-cost ratio of long-term treatment with lithium. Progress in this area would serve to shed light on the best balance between efficacy and side effects in clinical settings. Future research efforts may lead to more grounded guidelines, which are greatly needed in child and adolescent psychiatry.
Contributors
Authors AA, AO, JC, MT, CM and SNG designed the study and wrote the protocol. Studies were identified and independently reviewed for eligibility by two authors (AA, FS) in a two-step based process. Data were extracted by two authors (AA, FS) and supervised by a third author (SNG) using an ad-hoc developed data extraction spreadsheet. Authors AA, PO, AO, MA, AC and SNG wrote the first draft of the manuscript. Our manuscript has been approved by all authors.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests
Dr. Amerio, Dr. Ossola, Dr. Scagnelli, Dr. Odone, Dr. Allinovi, Dr. Cavalli, Dr. Iacopelli, Dr. Tonna, and Dr. Marchesi report no conflicts of interest. Dr. Ghaemi has provided research consulting to Sunovion in the past year, and is employed by Novartis Institutes for Biomedical Research. Neither he nor his family hold equity positions in pharmaceutical corporations.
Acknowledgements
The authors would like to thank Luigi Rolli for his help and support.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the online version, at doi: https://doi.org/10.1016/j.eurpsy.2018.07.012.
Comments
No Comments have been published for this article.