Special populations: Treatment resistant mood disorders

BACKGROUND

For the purposes of this manuscript, ‘mood disorders’ is defined as both depression and bipolar disorder. Mood disorders are highly prevalent throughout the United States. Unfortunately, the high rates of relapse, recurrence, and treatment failure leads to treatment resistance despite adequate medication trials. There are varying definitions of treatment resistance, but in general it is the lack of response to trials with two different medication classes of adequate dose and duration for either depression or mania.¹² There is a lack of strong evidence in treatment guidelines for the management of resistant mood disorders overall, and even less information for non-adult populations. This article will review the available literature and treatment recommendations for treatment resistant mood disorders in special populations including: geriatrics, children and adolescents, pregnancy, and comorbid substance use disorders.

GERIATRICS

In patients above the age of 60, the combined prevalence of major depression, dysthymic disorder, and “minor” depression has been estimated to be 25%.³ Major depressive disorder in nursing home residents has been reported to be between 14% and 42%.³ While elderly patients represent a minority of the United States population living with depression, they account for 18% of all suicide deaths.⁴ Prevalence of depression among this population may be higher as it may be underdiagnosed or misdiagnosed. Even further, the elderly population is less responsive to conventional treatments with approximately one-third responding to antidepressants with adequate dose and duration.⁵ This compares to response rates of 50 to 75% in clinical trials consisting primarily of non-elderly adults.³ Clinicians are presented with a difficult task of managing these patients in the setting of limited evidence, age-related pharmacodynamic and pharmacokinetic changes, increased susceptibility to side effects, polypharmacy, and comorbid medical illness. Due to the lack of evidence, the available guidelines are based primarily on expert opinion and do not address refractory patients.³⁶

Available evidence suggests anxiety and comorbid medical illness predispose elderly patients to treatment refractory depression resulting in prolonged time to remission and reduced remission rates. Individuals with anxiety are hypothesized to have increased sensitivity to side effects, poor adherence, and present with a more severe subtype of depression, but the mechanism is not well understood.⁷ Comorbid medical illnesses not only have been shown to negatively impact success with antidepressant augmentation with agents such as bupropion and nortriptyline but also maintenance outcomes with antidepressants.⁷ It has been theorized that medical illnesses affect the clinical utility of pharmacotherapy in depression by promoting pharmacodynamic and structural changes to the central nervous system.⁷ Executive dysfunction also has been suggested to lead to treatment refractory depression in elderly patients, but the effect on antidepressant response rates is unclear.⁷

Although the evidence in the literature for the treatment of resistant mood disorders in the elderly population is limited, there are some available data regarding both augmentation and switch treatment strategies for treatment resistant depression. These studies will be discussed below, but are limited due to low external validity, open label design, and small sample sizes. The studies also vary regarding definitions of response rate and remission (Table 1).

Table 1: Definitions of response and remission used in the studies of treatment resistant depression in the older population

View Fullscreen

Whyte et al. completed a post-hoc analysis of depressed elderly patients who received either augmentation (bupropion sustained release, nortriptyline, or lithium) or switch therapy (venlafaxine extended release) after failure of treatment with paroxetine and interpersonal psychotherapy. Overall, 60% of patients (32/53) responded to some form of augmentation (first augmentation trial: 24/53; second trial: 5/16; third trial: 3/7 responded). The average time to response after starting the first augmentation trial was 6 weeks. Twelve patients chose to switch to venlafaxine after failure to improve on paroxetine plus psychotherapy with or without augmentation in a 12 week open-label study. Three patients completed trials with paroxetine and all three augmentation strategies. Overall, 42% of patients (5/12) responded to venlafaxine during the 12 weeks, but the rate of response decreased with each subsequent treatment [first augmentation trial: 4/6 (67%); second trial: 1/3 (33%); third trial: 0/3 (0%) responded].⁴

Aripiprazole augmentation was evaluated in a 12 week open-label study of depressed patients 65 years and older who partially responded to sequential selective serotonin reuptake inhibitor (SSRI) and serotonin norepinephrine reuptake inhibitor (SNRI) pharmacotherapy. Aripiprazole augmentation was completed by 79% of the patients (19/24) with 50% achieving remission (12/24) and continuing treatment for a median of 27.6 weeks without any relapses. In terms of tolerability, akathisia occurred in 25% of patients (6/24), which was mild and transient except for one case leading to discontinuation of treatment.⁸

In an 8 week randomized, single-blind comparison of venlafaxine and paroxetine in elderly patients who failed two previous adequate trials of antidepressants, 60% (9/15) of patients receiving venlafaxine achieved remission compared to 33% (5/15) of patients receiving paroxetine. Response was achieved by 20% of patients (3/20) in each group. Venlafaxine resulted in a statistically significantly greater mean reduction from baseline to endpoint in the Clinical Global Impression Scale (–3.5 vs. –2.3; p<0.05) and Hamilton Rating Scale for Depression (–19.1 vs. –12.5; p<0.05). Overall, both treatments were well tolerated.⁹

Sequential treatment was evaluated in a two phase study. In the first phase, depressed patients 60 years and older received nortriptyline or venlafaxine in a 12 week double-blind, randomized, parallel-group trial. Remission was achieved in 32% of patients (26/81). Patients who failed to respond completed a 3 year, open-label, follow-up phase with lithium augmentation, phenelzine or tricyclic antidepressant (nortriptyline or clomipramine) switch therapy, or electroconvulsive therapy (ECT) per the Dutch consensus guidelines. Fifty-five patients entered the follow-up phase, but only 58% of patients (32/55) completed the three years. A majority of patients required more than one treatment (51 total treatments; 22 lithium augmentation, 13 nortriptyline, 2 clomipramine, 8 phenelzine, 5 ECT, and 1 selective serotonin reuptake inhibitor treatments). From overall pooled data, 96.3% of patients (78/81) achieved response and 85% (68/81) achieved remission in an average of 27.3 weeks with 2.3 treatment steps, including first phase treatment. Relapse occurred in 31% of patients (21/68) who achieved remission. Failure to achieve remission was associated with baseline characteristics of higher MADRS score, longer mean duration of depression, presence of a personality disorder, and increased concomitant medications to treat comorbid physical illnessess.¹⁰

For bipolar disorder, the prevalence ranges from 0.1–0.4% in patients 65 years and older, but the occurrence of late-onset bipolar disorder with the first episode of mania in this patient population is rare. These patients should be thoroughly evaluated for other underlying psychiatric, neurologic, medical, or medication-induced causes. For those who do develop late-onset bipolar disorder, there are higher rates of treatment refractoriness and psychotic features.⁶¹¹ According to the American Psychiatric Association practice guidelines for the treatment of patients with bipolar disorder, there are several factors to consider when treating bipolar disorder in this population. Geriatric patients require lower medication doses due to decreased volume of distribution, metabolism, and clearance. In addition, these patients are more sensitive to side effects, especially cognitive impairment, orthostatic hypotension, extrapyramidal symptoms, and tardive dyskinesia. Despite these considerations for treatment, there is a lack of evidence in the literature regarding treatment recommendations specifically targeted at geriatric patients with treatment resistant bipolar disorder.⁶

CHILDREN AND ADOLESCENTS

The child and adolescent population is a vulnerable population making it difficult to complete treatment studies. Therefore, identification of effective treatments for psychiatric disorders is still highly desired. Bipolar disorder affects approximately 1% of children and adolescents, and it is associated with psychotic symptoms, mixed mania or rapid cycling, and labile and erratic changes in mood, energy, and behavior.⁶¹² It occurs equally in both genders, but early-onset cases prior to the age of 13 years predominate in males.¹⁰ The course of illness is often more prolonged with treatment resistance and less inter-episode recovery is noted.⁶¹²¹³ Major depressive disorder (MDD) affects approximately 2% of children and 4–8% of adolescents, and it occurs in twice as many females compared to males during adolescence. MDD in adolescents is associated with more mood lability, irritability, temper tantrums, somatic complaints, and social withdrawal in addition to less melancholic symptoms, delusions, and suicide attempts compared to adults.¹⁴ For both bipolar disorder and MDD, comorbid diagnoses, such as anxiety disorders, conduct disorder, attention-deficit/hyperactivity disorder, and substance use disorders, are common in this population.³⁶¹³¹⁴

The American Academy of Child and Adolescent Psychiatry practice parameters for the treatment of bipolar disorder are mainly based on literature from adult populations since child and adolescent studies are limited to primarily open-label trials, case reports, and retrospective chart reviews. Lithium is FDA approved for acute mania and maintenance therapy in bipolar disorder patients 12 years and older, but valproate and carbamazepine, are also utilized.¹³ For the second generation antipsychotics, aripiprazole, olanzapine (ages 13–17), quetiapine, and risperidone are approved for the treatment of acute manic or mixed episodes in bipolar I disorder patients as monotherapy ages 10–17.¹⁵ Treatment-resistant bipolar disorder is not specifically addressed, but it is noted clozapine and ECT should be reserved for this subset of the population.¹³ To better clarify treatment-resistant mania in this population, Scheffer et al. developed and implemented a three-step algorithm in an open-label trial of treatment-resistant pediatric bipolar disorder patients aged 6–17 years with a Young Mania Rating Scale (YMRS) score of > 20 and at least two prior medication treatments (n=120). If symptoms persisted after discontinuing any destabilizing medications (e.g., antidepressants, stimulants, gamma aminobutyric acid agonists), optimizing anti-manic therapy (including lithium, valproic acid, and second-generations antipsychotics), and limiting treatment to 1 or 2 mood stabilizers, treatment-resistant illness was diagnosed. Overall, 75.8% of patients achieved remission (YMRS ≤ 12) after 6 months with 74.2% of patients achieving a greater than 50% reduction in their YMRS from baseline. Therefore, 24.2% of the population was diagnosed with true treatment-resistant mania.¹²

For the treatment of depression, the American Academy of Child and Adolescent Psychiatry practice parameters recommend antidepressants as monotherapy or in combination with psychotherapy for moderate to severe depression.¹⁴ Fluoxetine is the only FDA approved antidepressant for depression in children and adolescents. Treatment recommendations are limited due to a lack of positive treatment response and high placebo rates found in studies, and there are no specific guidelines for treatment-resistant depression. Strategies utilized in adults can be considered, such as optimizing current medication, switching to another medication, augmentation, combination therapy, or ECT.¹⁴¹⁶ The Treatment of Resistant Depression in Adolescents (TORDIA) study evaluated treatment options after failure of an 8 week trial with a SSRI in adolescents aged 12–18 years with MDD (n=334). After 12 weeks of treatment with a different SSRI, venlafaxine, SSRI plus cognitive behavioral therapy (CBT), or venlafaxine plus CBT, a higher treatment response rate was found with either combination therapy (55%) compared to monotherapy with another antidepressant (41%).¹⁷ Since CBT is expensive and not always available, Asarnow, et al. attempted to define predictors of treatment response in order to match patients with the optimal treatment strategy. The combination of CBT with an antidepressant may be more advantageous to utilize in adolescents with comorbid disorders (e.g., anxiety, attention-deficit/hyperactivity), no substance abuse history, and lower levels of hopelessness.¹⁸

PREGNANCY

Women are twice as likely to have depression compared to men with peak onset around childbearing years. Up to 13% of women living in the United States suffer from major depression, and it is estimated that 10–15% of women suffer from symptoms of depression during pregnancy or first year postpartum.¹⁹²⁰ Depression during pregnancy impacts obstetrical, fetal, and infantile outcomes. It has been shown to be associated with more sick days, physician visits, and pregnancy complications for the depressed mother, decreased initiation of breastfeeding and premature delivery for the infant, and decreased emotional bonding between the mother and infant.²⁰²¹²² Suicidal ideation during pregnancy ranges from 5–14%, and suicide accounts for up to 20% of postpartum deaths due to increased suicidality during the perinatal period.²³ The rate of treatment-resistant depression is unknown during pregnancy. Hendrick, et al. explored the differences of treatment response in patients with depression related and unrelated to childbearing. Postpartum women were more likely to present with anxiety symptoms, more severe cases of depression, a longer time to treatment response, and prescriptions for more than one antidepressant. By the end of 3 weeks of treatment, 75% of patients with depression unrelated to childbirth responded compared to only 33% of depressed postpartum patients.²⁴ Pregnancy has not been shown to increase the number of mood episodes in terms of bipolar disorder, but recurrence often occurs in this population if mood stabilizer therapy is abruptly discontinued. In addition, the postpartum relapse rate of bipolar disorder is as high as 50%.⁶

A full review of the use of mood stabilizers and antidepressants in pregnancy is beyond the scope of this article. However, due to limited evidence, the management of treatment resistant major depressive or bipolar disorders is the same as non-pregnant adults, but with additional considerations to minimize fetal harm. Treatment options include continuation of medication or discontinuation of medications prior to conception, at the beginning of pregnancy, or only for the first trimester.³⁶ In a systematic review, Udechuku, et al. found antidepressants to be relatively safe during pregnancy with the exception of paroxetine.²⁵ (Note: some authors have negated this finding, please refer to Mental Health Clinician of Psychotropic and treatment of the child-bearing woman Aug 2013 Vol.3, Iss. 2). Some concerns with antidepressants include higher preterm birth, neonatal adaptation difficulties, and congenital cardiac malformations.²⁵ In terms of the mood stabilizers, Galbally, et al. found lithium, carbamazepine, lamotrigine, and valproic acid to be associated with teratogenicity, with valproic acid (especially doses above 1000 mg) having the highest risk.²⁶ Some mood stabilizers, such as carbamazepine, affect folic acid metabolism but the risk of malformations may be mitigated if folic acid 5 mg is supplemented one month prior to conception and throughout the first trimester of pregnancy.²⁶

A population-based, case-control study evaluating congenital abnormality risk of antiepileptic exposure in utero (i.e., carbamazepine, phenobarbital, phenytoin, primidone), which was limited to epilepsy data, found the risk was reduced but not eliminated by folic acid supplementation. The odds ratio of congenital abnormalities with carbamazepine and folic acid supplementation was 1.31 (95% CI 0.43–4.02) compared to 3.34 (95% CI 1.49–7.49) without folic acid supplementation.²⁷ In this study the dose of folic acid was only available for 600 women (of the 600 women 22.5% took 3 mg, 68.8% took 6 mg, and 8.9% took 9 mg/day).²⁷ For antipsychotics, the fetal risk cannot be determined due to limited data, but high potency antipsychotics are preferred due to less anticholinergic, antihistaminergic, and hypotensive effects.⁶ It is unclear whether first or second generation antipsychotics are more harmful to the fetus. However, second generation antipsychotics have been associated with increased risk of gestational metabolic complications and higher mean birth weight compared to first generation antipsychotics.²⁸

Due to lack of evidence regarding treatment in this population, especially for treatment-resistant disorders, the lowest effective dose should be utilized. Decreased response to pharmacotherapy and the occurrence of resistance during pregnancy may be the result of pharmacokinetic changes, such as increased volume of distribution, decreased concentration of plasma proteins, increased hepatic function, and hormone-induced changes in metabolic enzymes. Additionally, polypharmacy may have additive effects on fetal development, so doses should be maximized prior to switching or adding another medication.²⁹

SUBSTANCE USE DISORDERS

Substance use disorders often co-occur with mood disorders. When comparing Axis I disorders per the DSM-IV-TR criteria, bipolar disorder has been found to have the highest life-time rates of alcohol and drug use disorders (46% and 41%, respectively) compared to the general population (14% and 6%, respectively).⁶²⁹³⁰ The presence of substance use disorders has been shown to negatively impact the course and outcome of treatment for mood disorders, including bipolar disorder and major depression. Patients with substance use disorders and co-occurring bipolar or major depressive disorder have been found to have an earlier age of onset, more severe subtypes with rapid cycling or mixed episodes, higher rates of hospitalizations, increased rates of suicidal ideations and suicide attempts, and decreased medication adherence when concurrently using substances.³⁶³⁰³¹

According to the American Psychiatric Association practice guidelines for the treatment of patients with depression, bipolar disorder, and substance use disorders, there are many factors requiring consideration when treating patients with concurrent substance use and mood disorders. If a patient is still using substances, the medication regimen needs to be evaluated for synergy of actions (e.g., benzodiazepines and alcohol) in addition to drug interactions (e.g., stimulants and MAOIs; tobacco smoke and antipsychotics). Some medications utilized to treat mood disorders, such as benzodiazepines and stimulants, have a potential for abuse and dependence leading to a desire to avoid their use in this patient population. The effects of substance use on medication metabolism must also be considered. For example, dehydration secondary to alcohol use can predispose a patient to lithium toxicity. Additionally, hepatic dysfunction secondary to alcohol use or intravenous substance use can lead to elevated levels of valproic acid or carbamazepine. Two final factors to consider when treating concurrent substance use and mood disorders are risk of intentional and unintentional overdose in addition to the potential for non-adherence if a patient is intoxicated or withdrawing from substance use.³⁶³²

Despite these considerations for treatment, there is a lack of evidence in the literature regarding treatment recommendations specifically targeted at patients with treatment resistant mood disorders and concurrent substance use disorders. Additionally, the prevalence of treatment resistant mood disorders in patients with substance abuse disorders has not been determined. In a study by Pacchiarotti et al., patients whose substance use preceded the onset of bipolar disorder were found to have a less severe course of illness with fewer total lifetime hypomanic and depressive episodes compared to those whose bipolar disorder preceded the substance use. This does not mean patients whose bipolar disorder preceded the onset of substance use will be treatment resistant.³¹

CONCLUSION

In summary, the literature on the treatment of resistant mood disorders is limited and is even rarer in special patient populations. While geriatric, children and adolescents, pregnant women, and patients with comorbid substance use disorder represent a minority of the U.S. population living with mood disorders, they are likely to have more dire consequences with failed treatment. The elderly are less likely to respond to usual treatment of mood disorders. Those that do respond will have a prolonged time to response and remission, especially in those with comorbid anxiety and medical illness. In caring for an elderly patient, it is important to consider pharmacokinetic changes, use the lowest effective dose, and limit polypharmacy. In children and adolescents, the majority of the data are derived from open-label and/or adult studies. While there are no specific treatment recommendations for treatment resistance in this patient population, clozapine and ECT may be viable options. In treatment resistant depression in children and adolescents, it was found that antidepressants in combination with CBT produced a higher response rate compared to monotherapy with antidepressants. In pregnant women, it is important to use the lowest effective dose and limit polypharmacy to prevent fetal harm. Additionally, pharmacokinetic changes throughout pregnancy may warrant a change in dose. More frequent monitoring is recommended in this population. There are no recommendations specifically for treatment of patients with treatment resistant mood disorders with comorbid substance use; however, it is important to promote abstinence. Many substances of abuse may interact with medications leading to treatment resistance and harm. Future studies with standard definitions of treatment resistance, response, remission, and treatment failure are needed in these populations to better guide clinicians on their treatment.