Psychopharmacology in the Pediatric Oncology and Bone Marrow Transplant Units: Antidepressant Treatment

Introduction

Pediatric cancer is undoubtedly one of the most difficult life events for a child and his family (Bruce 2006; Long and Marsland 2011). Despite improving survival rates, the diagnosis and treatment of childhood cancer still impose a considerable physical and mental burden on patients. Commonly identified stressors include physical side effects, such as fatigue, nausea, and pain, as well as more social stressors such as appearance changes, missing home due to hospital confinement, and limited social support (Sharp et al. 2017). Many young cancer patients experience psychiatric comorbidity; depression, anxiety, and adjustment symptoms most commonly precipitate psychiatric referrals in this patient population (Sohn et al. 2017; Shim et al. 2018; Samsel et al. 2019). In children suffering from cancer, the reported rates of depression and anxiety disorders vary between of 7% and 32%, probably depending on the methodology of individual studies.

The diagnosis of major depressive disorder in children is complicated by many factors, including variable age-dependent presentations of depression; varying perceptions of the child's symptoms between caregivers (parents and clinicians); and the individual trajectory of any patient's treatment course (diagnosis, induction therapy, relapse, and end-of-life care are examples). Of pediatric cancer patients referred for psychiatric evaluation, some meet criteria for major depression or anxiety disorders, while others exhibit dysthymia, depressed mood, or an adjustment disorder (Kersun and Elia 2007). Premorbid psychiatric illness and parental distress contribute to the psychiatric morbidity seen in pediatric oncological patients (Okado et al. 2014).

Differentiating between the clinically significant symptoms of a mood disorder, foreseeable adjustment difficulties, and disease/treatment-related effects can be challenging due to frequent overlap of cancer symptomatology and diagnostic criteria for depression (Portteus et al. 2006; Phipps et al. 2012). Anhedonia, weight or appetite changes, insomnia or hypersomnia, psychomotor agitation or retardation, fatigue or loss of energy, and diminished ability to concentrate are frequent in children with cancer; determining an oncologic versus a psychiatric etiology can be difficult. These challenges require a team approach that includes effective communication between oncologists, nurses, mental health and child professionals, and social support teams.

Attention to the mental health needs of these children has evolved into a specialty, psycho-oncology, which is increasingly being recognized as an essential component in the already complex care of children with cancer and their families (Datta et al. 2019).

The detrimental impact of depression and anxiety on various outcomes in cancer patients has been well established (Shim et al. 2018). Depression amplifies medical disease-related disability, impairs quality of life, and has been repeatedly shown to be associated with poor medical outcomes and early death in children who suffer from life-threatening medical conditions (Shemesh et al. 2002). Untreated depression may also interfere with immune system functioning (Apter et al. 2003), potentially influencing the incidence of infectious complications in these already immunocompromised patients.

Furthermore, depression and anxiety are associated with treatment nonadherence (Kennard et al. 2004; Portteus et al. 2006; Kurtz and Abrams 2010; Arrieta et al. 2013; Caruso et al. 2017; Shim et al. 2018), which can have devastating ramifications on treatment efficacy and the risk of disease relapse (Lauer 2015). Nonadherence to treatment regimens in critically ill patients is addressed in the Diagnostic and Statistical Manual of Mental Disorders, 5th ed. (American Psychiatric Association 2013).

Psychopharmacological management serves as an important element of a comprehensive multidisciplinary approach to the care of pediatric cancer patients (Pao et al. 2006; Samsel et al. 2019). Selective serotonin reuptake inhibitors (SSRIs) are the most prescribed antidepressants in children, despite diverse side effects, including the 2004 black box warning issued by the FDA. Prescription of SSRIs is common practice among pediatric oncologists, often in the absence of (or before) consultation with mental health professionals (Kersun and Kazak 2006; Portteus et al. 2006; Pao et al. 2009; Phipps et al. 2012; Mavrides and Pao 2014).

Despite widespread use, there are limited published studies exploring the use of psychopharmacological treatments for depression and anxiety in children with cancer. Most studies focus on the use of SSRIs, with little data on other antidepressants, such as mirtazapine, which holds a unique pharmacological profile (Gothelf et al. 2005; DeJong and Fombonne 2007; Kersun and Elia 2007).

Despite the absence of definitive clinical data regarding the efficacy and tolerability of antidepressants in children (Steele et al. 2015), this class of drugs seems to be beneficial for the treatment of symptoms of depression and anxiety, such as depressed mood, loss of appetite, fatigue, anxiety, and cognitive impairment, as well as for other associated symptoms such as pain, gastrointestinal tract distress, wasting, hot flashes, vasomotor symptoms, sleep problems, and pruritus in children and in adults (Pao et al. 2006; Zaini et al. 2018; Belinskaia et al. 2019).

In the absence of large, controlled studies, guidelines for the use of psychopharmacological treatment of children with cancer and comorbid psychiatric diseases have yet to be established. The goal of this retrospective study is to describe our experience with the administration of antidepressants in a large pediatric Hematology–Oncology center. We evaluated our experience with the use of antidepressants, including efficacy and tolerability, and assessed potential drug interactions between the psychopharmacological treatments and medications prescribed for the treatment of malignant disease and as supportive care.

Methods

Results

Clinical Global Impressions scores

The mean CGI-S score at first assessment of the entire cohort was 5 ± 1.07 (4.7 ± 0.9 and 5.4 ± 1.1 in the SSRIs and mirtazapine recipients, respectively, with no between-group statistically significant difference). Following the initiation of treatment, patients were reassessed at a median of 31 days, (quartile percentiles: 25%: 13 days, 50%: 25 days, 75%: 41 days). The mean CGI-S at reassessment (3.4 ± 1.3), was significantly improved as compared with the first pretreatment time point (F _2,28 = 156, p < 0.001; see Fig. 1) with no statistically significant difference between SSRIs and mirtazapine-treated patients (3.1 ± 1.2 vs. 3.9 ± 1.4, p > 0.15; see Fig. 2). Univariate ANOVA did not reveal any significant effect of reassessment intervals on the model ( F _2,28 = 0.46, p > 0.63).

OPEN IN VIEWER

FIG. 1.

CGI-S Scores of pediatric oncologic patients (n = 32) at first and second assessments, that is, before and after the antidepressant treatment, respectively. Antidepressant treatment significantly (F _2,28 = 156, p < 0.001) reduced CGI-S scores. There were no statistically significant differences in CGI-S scores between the SSRI's and mirtazapine groups (3.1 ± 1.2 vs. 3.9 ± 1.4, p > 0.15). The effect of number of days between assessments between groups was not significant (F _2,28 = = 0.46, p > 0.63). CGI-S, Clinical Global Impressions-Severity.

OPEN IN VIEWER

FIG. 2.

CGI-S scores of pediatric oncologic patients (n = 32) treated with SSRI's versus mirtazapine. No statistically significant difference was found between CGI-I scores of SSRI's and mirtazapine groups (2.6 ± 1.2 vs. 2.2 + 1.2, p > 0.40). The effect of number of days between assessments was not significant (F _2,28 = 0.46, p > 0.63). CGI-I, Clinical Global Impressions-Improvement.

Twenty-five of 32 patients (78%) showed improvement in the CGI-I score—6 were “very much improved,” 13 “much improved,” 6 “minimally improved,” and 6 showed neither improvement nor aggravation. One patient, who discontinued medication due to side effects, was not included in the analysis. There was no statistically significant difference in CGI-I score between the two groups of antidepressants (2.6 ± 1.2 vs. 2.2 ± 1.2, p > 0.40).

Discussion

The diagnosis and treatment of childhood cancer impose considerable physical and psychological burdens on patients; psychopharmacological interventions, together with other nonpharmacological treatments, are an important element of the interdisciplinary treatment approach of these children (Samsel et al. 2019). The use of antidepressants in the treatment of anxiety and depression in children with cancer has increased (Pao et al. 2006; Kurtz and Abrams 2010), but there are only few small studies on the treatment of these challenging patients (Gothelf et al. 2005; DeJong and Fombonne 2007; Kersun and Elia 2007; Steele et al. 2015).

Our retrospective study reviews the use of two groups of antidepressants prescribed by liaison psychiatrists at the Hematology–Oncology Department of the Schneider Children's Medical Center in Israel. Our main finding shows that the use of SSRIs and mirtazapine in this population is well tolerated, in a wide age range, with benefits that outweigh adverse reactions. We demonstrated a combined clinical and pharmacological approach through use of a multidisciplinary team consisting of a child psychiatrist, clinical pharmacist, and pediatric oncologist.

The diagnosis of psychiatric disorders in cancer patients can be confusing and sometimes even misleading. There is considerable and challenging overlap between the presentation of psychiatric illness in general and mood disorders in particular and the symptomatology that follows the diagnosis of cancer, as well as the adverse effects of its treatment (Portteus et al. 2006; Phipps et al. 2012).

Such symptoms as anhedonia, weight or appetite loss, insomnia, psychomotor agitation or retardation, fatigue or loss of energy, and diminished ability to concentrate are frequent in children with cancer. Concomitant somatic symptoms present a special challenge to the mental health practitioner, as they may be signs of either the primary disease or of an underlying mental illness. Thus, standard depression criteria and questionnaires may be of more limited utility in medically ill patients (Cohen-Cole and Stoudemire 1987; Canning and Kelleher 1994; Knapp and Harris 1998), while a symptom-oriented approach may be more appropriate (Fleishman 2004; Pao et al. 2006; Peled et al. 2020). Clearly, a “one-size-fits-all” approach is inadequate both in the diagnostic and treatment contexts.

Nearly all our patients (94%) reported “depressed mood” as a symptom that led to the use of antidepressants, even though half of our patients did not fulfill the DSM-5 criteria of a major depressive disorder. It has been reported that depression in cancer patients often does not reach cutoff thresholds for “clinically significant depression,” but nonetheless, depressive symptoms negatively affect patients' quality of life (Mavrides and Pao 2014).

The most frequent psychiatric diagnosis in this study was major depressive disorder (47%), followed by anxiety and medication-induced psychiatric disorders (19% each), findings similar to those reported in adults with cancer (Cullivan et al. 1998). At the time of psychiatric evaluation, 69% of the patients were receiving different nonpsychiatric medications likely to induce neuropsychiatric adverse events, although all the medication-induced psychiatric disorders in our cohort were attributable to corticosteroid usage. The use of systemic corticosteroids can induce various psychiatric changes, ranging from behavioral and mood changes to psychotic symptoms (Caplan et al. 2017). It is noteworthy that corticosteroid-associated psychopathology can occur during therapy, during steroid taper, or following discontinuation of the drug (Drozdowicz and Bostwick 2014).

The beneficial therapeutic effect of antidepressants in our patients was similar in SSRIs and mirtazapine-treated patients. Adverse effects of treatment were infrequent and mild, but of sufficient magnitude to lead two patients to stop treatment. Among our mirtazapine-treated patients, elevated liver enzymes and jaundice were noted in one patient, and causality was not demonstrated.

The investigated antidepressants in this study differ in their pharmacological profiles. SSRIs tend to have similar side-effect profiles, however, certain SSRIs may be more likely to cause specific side effects (Nierenberg et al. 2008). Common SSRI side effects include sexual dysfunction, drowsiness, weight gain, insomnia, agitation, anxiety, dizziness, orthostatic hypotension, headache, and dry mouth. The mild side effects reported by the three patients treated within the relatively narrow range of SSRI dosages prescribed are compatible with the side-effect profile of SSRIs.

Among antidepressant medications, mirtazapine possesses a unique pharmacological profile, potentially rendering it effective in treating multiple symptoms such as depression, anxiety, sleep disorders, cachexia, and neuropathic pain (Davis et al. 2002; Lovell and Ward 2011; Zaini et al. 2018). Mirtazapine, a noradrenergic and specific serotonergic receptor antagonist, has proved effective in the treatment of depression among patients with cancer (Lovell and Ward 2011; Economos et al. 2020).

Although data regarding the use of mirtazapine in children with cancer are sparse, the drug was shown to be safe and well tolerated in children (Gray et al. 2018). Mirtazapine presents a variety of side effects, with drowsiness and dizziness being the most common (Economos et al. 2020). Moreover, some of the commonly described “adverse” effects of mirtazapine, such as sedation, appetite stimulation, and weight increase, are exclusive to mirtazapine and have potential palliative benefits.

Additionally, recent studies describe mirtazapine's utility in relieving refractory pruritus (Demierre and Taverna 2006; Sheen et al. 2008; Lee et al. 2016; Bromberg et al. 2020; Fawaz et al. 2021), making it the go-to medication for patients with graft versus host disease, who require treatment for concurrent depression (Davis et al. 2002; Lovell and Ward 2011; Zaini et al. 2018). For these and other reasons, mirtazapine was prescribed to a particularly young patient of ours (2 years old); the medication was well tolerated and proved to be beneficial.

The precipitation of bleeding due to serotonin depletion by SSRIs has been described and referred before in postchemotherapy and poststem cell transplantation thrombocytopenic patients (Tseng et al. 2010, 2013); None of the patients in our cohort experienced SSRI-induced thrombocytopathy and bleeding. Nevertheless, we have encountered two cases of laboratory-confirmed SSRI-induced inhibited platelet aggregation and critical bleeding in the past, which have raised our awareness to this adverse event. Therefore, we preferred mirtazapine in patients who were at risk for bleeding (Clinical Pharmacology 2021).

Children with cancer often receive many different medications, both for treatment of the primary oncological disease and for treatment or prevention of infectious and noninfectious complications of therapy. Drug–drug interactions between concomitant medications and antidepressants are a cause of concern. Antidepressants must be chosen not only based on potential efficacy and side-effect profile, but with careful attention to the proposed agent's toxicity and interactions with the medications that the patient already receives. Table 5 illustrates a profile of drug–drug interactions that should assist cautious choice of antidepressants in children with cancer.

All SSRIs may inhibit one or more cytochrome P450 drug-metabolizing enzymes, causing pharmacokinetic interactions with many drugs. Most SSRIs are weak inhibitors of CYP450 enzymes and unlikely to cause clinically relevant pharmacokinetic interactions. In contrast, fluoxetine is a potent inhibitor of CYP2D6 and a moderate inhibitor of CYP2C19, suggesting that its use may be inappropriate in children with cancer and explaining its infrequent use in our cohort. Fluvoxamine was not prescribed in our cohort, as it is a potent inhibitor of CYP1A2, with moderate effects on CYP2C9, CYP2C19, and CYP3A4. Among SSRIs, escitalopram and sertraline are the drugs of choice for situations in which drug–drug interactions are a concern (Preskorn 1997; Labbate et al. 2010).

The narrow therapeutic window and wide inter- and intrapatients' variability of many of these drugs require initial and ongoing consultation with a clinical pharmacologist during the treatment, taking into consideration the list of concomitant medications involved in the course of the oncological therapy.

We also note that antidepressants are often prescribed by pediatric oncologists without consultation with a liaison psychiatrist or a clinical pharmacologist (Kersun and Kazak 2006; Portteus et al. 2006; Pao et al. 2009; Phipps et al. 2012; Mavrides and Pao 2014). Forty-one percent of the patients was already receiving one or more psychotropic medications at the time of psychiatric evaluation, highlighting the need for access to a specialized psychiatric consultation. Rigorous follow-up by mental-health professionals is important to monitor the efficacy of treatment and to permit timely change of treatment as dictated by the dynamic nature of the treatment trajectory of children with cancer.

The major limitations of this study are its open, retrospective nature and its small sample size. The treatment outcome was assessed based on the clinical global impressions of the psychiatrist at the time of examination, in an unblinded process (the evaluator knew both that medication had been given and which agent had been used, potentially affecting his or her assessment of the magnitude of improvement). Standardized scales for psychiatric evaluation were not used. Finally, the current study presents the practice patterns in one treatment center; its results require confirmation in other sites. We hope that this report will encourage large-scale, multisite, prospective studies of the use of antidepressants in Pediatric Oncology practice.

Conclusions

The antidepressants (fluoxetine, escitalopram, sertraline, and mirtazapine) used in this study population were found to be well tolerated and beneficial, suggesting a role in ameliorating acute psychiatric symptoms in children and adolescents with cancer. Clinicians may elect to utilize some secondary effects of mirtazapine for palliation of symptoms associated with cancer and its treatment in young patients.

Clinical Significance

Our findings have important implications for pediatric oncologists; pediatricians providing supportive care for children with cancer in both outpatient and inpatient settings; and child psychiatrists. Furthermore, our study has the benefit of including a wide age range (2–21), a unique strength among published works in the field. Our study emphasizes the importance of psychiatric and pharmacological evaluations of children with cancer before starting antidepressant therapy, to provide safe and efficacious management and monitoring. The antidepressants used in this study population were found to be beneficial and well tolerated. Prospective clinical trials are needed to explore the relative benefits of different specific antidepressants and to validate our findings. Such studies will facilitate the elaboration of specific guidelines regarding the use of antidepressants for children with malignant diseases.