Safety of Attention-Deficit/Hyperactivity Disorder Medications in Children: An Intensive Pharmacosurveillance Monitoring Study

Introduction

A ttention-deficit/hyperactivity disorder (ADHD) is a disorder of neuropsychiatric development of children and adolescents characterized by inattention and impulsivity/hyperactivity, according to the Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV-TR) (American Psychiatric Association 2000). These symptoms are generally accompanied by depression, anxiety, oppositional defiant disorders (ODDs), and compulsive behaviors. Specifically, the DSM-IV-TR distinguishes three clinical types of ADHD: inattentive (ADHD-I), hyperactive (ADHD-H) and combined (ADHD-C).

The prevalence of ADHD is very heterogeneous and changes across countries. In fact, 7.8% of children and adolescents (4–17 years) living in the United States have ADHD (Centers for Disease Control and Prevention 2005), whereas in Italy, the prevalence of the syndrome is estimated at between 1% and 4% of children (5–12 years old) (Panei et al. 2004). The worldwide mean prevalence has recently been estimated to be 5% in children (Polanczy et al. 2007) and 3% among adults (Fayyad et al. 2007). In ∼80% of children with ADHD, the symptomatic features continue into adolescence and in 30–65% even into adulthood (Vallejo et al. 2009).

Treatment of ADHD is multimodal and includes medication management, behavioral interventions, and a combination of both approaches (Kaiser et al. 2008). Stimulant medications, such as methylphenidate (MPH), and nonstimulant medications, such as atomoxetine (ATX), are used for the pharmacological treatment of ADHD in children (National Institute for Health and Clinical Excellence 2006). A number of studies have demonstrated the efficacy of both drugs in patients with ADHD versus placebo (Cotton and Rothberg 1988; Klorman et al. 1990; Christman et al. 2004). Nevertheless, there is some concern about their long-term effectiveness (Jensen et al. 2007). Although relatively safe, both MPH and ATX have class-related warnings and contraindications and are associated with adverse effects that require consideration when prescribing (Wolraich et al. 2007). A number of trials involving ATX and MPH indicate that adverse effects are usually mild, and can be managed with medication withdrawal. These symptoms include neurological effects (headache, dizziness, insomnia, and seizures), psychiatric effects (mood/anxiety, tics, psychotic symptoms), and gastroenterological effects including poor appetite and overlapping with growth restriction (Graham et al. 2011). Nevertheless, such serious adverse effects as sudden cardiac death and suicidality require further investigation to allow a more precise understanding of these risks (Graham et al. 2011). Moreover, since their authorization, the use of ADHD drugs has given rise to ethical and legal issues (Foreman 2009).

Recent years have seen an exponential increase in ADHD diagnoses and in psychotropic drug prescriptions (Bonati and Clavenna 2005). A number of restrictive measures have been adopted to stop this trend. In 2007, following the reintroduction of MPH on the Italian market and the registration of ATX, the Italian drug agency (AIFA) introduced the ADHD National Register, an active tool with which to monitor the appropriateness of care, therapy efficacy, and adverse drug reactions (ADRs), and to prevent drug misuse. The ADHD National Register includes children enrolled by reference prescription centers. These centers, located throughout Italy, confirm the ADHD diagnosis and establish the pharmacological therapy. There are five reference prescription centers in the Campania Region, whereas the Regional Center of Pharmacosurveillance and Pharmacoepidemiology of the Second University of Naples is responsible for monitoring the risk/benefit ratio. Since its institution, one of the most important Regional Center objectives is to investigate drug safety in children. This aim is of paramount importance because children are not generally included in clinical trials and, therefore, drugs are prescribed based on data derived from trials performed with adults. Psychostimulants are an exception, because most of the subjects included in clinical trials are children; nevertheless, as clinical trials have time limits, these drugs are still under concern in terms of long-term tolerability. Despite the fact that in Italy spontaneous reporting of ADRs is mandatory, there is still a high frequency of under-reporting of ADRs in children. Therefore, as part of its postmarketing surveillance activities, the Regional Center of Pharmacosurveillance and Pharmacoepidemiology of the Second University of Naples started an intensive pharmacosurveillance monitoring program to increase the number of ADRs recorded in the Italian spontaneous reporting database, and to collect information on the safety profile of ATX and MPH in a Campania pediatric population with ADHD.

Methods

From September 2007 to October 2010, the Campania Center of Pharmacovigilance and Pharmacoepidemiology started an intensive pharmacosurveillance monitoring study of children with ADHD.

Results

From September 2007 to October 2010, 1841 children and adolescents were listed in the Italian ADHD Register, of which 918 were treated with ATX (49.8%) and 924 were treated with MPH (50.2%) (Table 1). There was a preponderance of: males (88.6%), the 10–13-year-old age group (40.3%), the ADHD C subtype (84.2%), and learning disability as the main comorbidity (44.8%). Moreover, in Italy, 18% of subjects were treated with another drug in addition to ATX or MPH.

During our study, we enrolled 76 subjects which represents 4.1% of the entire cohort listed in the National Register. As shown in Table 1, the clinical and demographic characteristics of our study population were largely consistent with those of subjects listed in the National Register (86.8% in Campania were male; 34.2% were 10–13 years old, 64.5% had a diagnosis of ADHD-C, and 48.7% had a learning disability). The main differences were related to the prevalence of subjects diagnosed with hyperactive subtype (25% in Campania vs. 4.4% in Italy overall) and the prevalence of those with a behavioral disorder as comorbidity (22.4% vs. 6.6%). Moreover, although in Italy as a whole, patients were evenly distributed in the two treatment groups, almost the entirety (68 subjects) of our study population was treated with ATX (89.4%).

In addition to ATX or MPH, data obtained from the National Register showed that 18% of Campania subjects were additionally treated with another drug. Among subjects enrolled in the Campania Region, 12 (15.8%) received at least one concomitant medication: valproate (6), risperidone (5), ethosuximide (1), alprazolam (1), growth hormone (GH) (1), levothyroxine (1), phenobarbital (1 child), niaprazine (1), carbamazepine (1), and topiramate (1) (data not shown). We assessed if those on concomitant medications had more ADRs than did those on ATX or MPH alone.

At the time of analysis, among our study population the median follow-up period was 289 days (∼9 months); 40 (52.6%) enrolled subjects stopped treatment and dropped out of the register. The reasons for dropping out were ADRs (27.5%); improvement of clinical condition (37.5%); perceived lack of efficacy (10%); and such other causes as patient decision, poor compliance, loss to follow up, or moving to another structure (25.0%) (Table 2).

Discussion

To our knowledge, this is the first intensive pharmacosurveillance monitoring study that used not only the ADHD National Register data, but also the Italian spontaneous reporting database to investigate the tolerability of ADHD medications in a pediatric population. Other authors (Didoni et al. 2011), in a recent study, demonstrated the importance of the national register for monitoring the safety and efficacy of drug therapy. Nevertheless, data from registers generally lack full information; therefore, first we designed an ad hoc form to collect more thorough data about ATX and MPH safety and introduced expert monitors who closely collaborated with clinicians.

Based on our findings, only 4% of subjects in the ADHD National Registry were enrolled from Campania. This can be explained by a distinctive local approach of preferring other therapies, such as behavioral therapy. Moreover, the limited number of reference prescription centers in Campania should be also taken in consideration. The distinct approach to ADHD in Campania is also confirmed by the different distribution of patients in the ATX and MPH groups. Whereas in the ADHD National Register a similar number of patients were treated with ATX or MPH, in Campania ∼90% of enrolled patients received ATX and only ∼10% received MPH.

Previous studies have shown that ATX and MPH have a similar efficacy (Cotton and Rothberg 1988; Klorman et al. 1990; Christman et al. 2004). On the contrary, other authors have reported that OROS MPH, a long-acting methylphenidate preparation (not marketed in Italy), had greater efficacy than ATX (Newcorn et al. 2008). Probably, the trend in MPH prescribing in Campania may reflect the fact that, as required by the AIFA-approved protocol, the first dose of MPH must be administered in a hospital and a number of Campania reference prescription centers lack a day hospital unit. The high percentage of patients treated with ATX in Campania gave us a unique opportunity to focus on the safety profile of this drug that still requires close monitoring because of its recent approval. In contrast, MPH was approved >50 years ago; therefore, its safety profile is well known and there was only a remote possibility of identifying rare or unexpected ADRs.

In addition to the pharmacological therapy, the demographic and clinical characteristics of Campania subjects were similar to those of other subjects enrolled in the National Register: Most of were male, between 10 and 13 years of age, had a diagnosis of ADHD-C, and had a learning disability as the main comorbidity. The prevalence of these characteristics in children and adolescents with ADHD varies across studies (Polanczyk et al. 2007; Skounti et al. 2007). However, it is generally agreed that male gender entails a greater risk of ADHD (Biederman et al. 2002). The prevalence in males may be related to under-identification and underdiagnosis of ADHD in females. Evidence suggests that females are more likely to have an adolescent inattentive disease, which is more easily overlooked than the disruptive component of ADHD that tends to affect males (Biederman et al. 2002).

In our study, valproate was the most frequently used add-on therapy, and it was administered to patients with epilepsy associated with ADHD. Parisi et al. (2010) recently reported a high frequency of association between epilepsy and ADHD, and suggested that there is bidirectional relationship between the two disorders.

Twenty-five (32.8%) patients from Campania reported at least one ATX- or MPH-induced ADR and five of these (20%) were on a concomitant psychiatric medication. Among all the ATX- or MPH-induced ADRs recorded in the Italian spontaneous reporting database, most occurred in patients from Campania, probably because of our intensive monitoring program.

In our study, the clinical and demographic characteristics were similar in subjects with and without ADRs. In most cases, ADRs were not serious and resolved after drug withdrawal. As expected, because of the unbalanced distribution of patients in the treatment groups, most ADRs were related to ATX treatment. The most common ADR was weight loss. In our study, we recorded this event in nine patients (eight on ATX, one on MPH1). In most cases, this symptom was mild-to-moderate and tended to subside over time. Only one case was clinically significant and led to discontinuation of therapy. We also found that weight loss was not related to days of therapy. Some patients experienced the most consistent weight loss after fewer days of therapy than other patients. This is in line with previous findings that both ATX and MPH may induce weight loss (Spencer et al. 2007; Faraone et al. 2008) especially during the initial months of treatment. Weight loss may be secondary to decreased appetite which in turn may be related to reported gastrointestinal symptoms such as nausea, vomiting, and/or upper abdominal pain. Alternatively, weight loss may occur consequent to loss of appetite probably resulting from temporary derangement of the central noradrenergic systems involved in hunger or satiety (Spencer et al. 2007).

All reported ADRs were expected, because they are listed in the summaries of product characteristics, except for a case of weight gain and a case of buccofacial dystonia, both related to ATX treatment. To our knowledge, ATX-induced weight gain has not been reported previously.

The one instance of weight gain was reported in an 8-year-old boy who gained 10 kg between the 6th and the 9th month of treatment with ATX 18 mg/day. He was not under treatment with other drugs. After exclusion of other possible causes such as increased food intake, the clinicians suspected that the weight gain was related to ATX. The Naranjo algorithm score was consistent with a possible causality link.

The case of dystonia was reported in a boy undergoing treatment with ATX 18 mg/day for 1 month. He had epilepsy and had been taking valproate for 3 months. Four days after he started cotreatment with risperidone 0.5 mg/day, he experienced buccofacial dystonia, which resolved after risperidone withdrawal. We believe this adverse reaction was related to an interaction between risperidone and ATX. Indeed, both ATX and risperidone are metabolized by the cytochrome P-450 2D6 isoform (CYP2D6) (Fang et al. 1999; Sauer et al. 2005). It is conceivable that ATX may have reduced the metabolism of risperidone and increased its plasma concentrations, thereby leading to dystonia. This possibility is substantiated by reports of extrapyramidal symptoms associated with ATX in combination with other drugs (Bond et al. 2007). Other possible explanations are that ATX induced an atypical effect, an increased level of ATX or its metabolites in the circulation because of the patient's poor metabolizer status or because of excess synaptic norepinephrine or dopamine (Bond et al. 2007). Although risperidone is not contraindicated in patients under treatment with ATX, clinicians should be aware of the possibility of a metabolic interaction with ATX that may lead to exacerbation of adverse extrapyramidal effects.

Another important ADR identified in this study was suicidal ideation. The patient experienced worsening of oppositional behavior and threatened to commit suicide 5 months after starting ATX 40 mg/day. He was not on concomitant medications. The clinician reduced the dose to 25 mg/day; the outcome was not reported. Suicidal ideation is a rare and extremely serious ATX adverse event that has been identified by spontaneous postmarketing alerts. Thus far, the risk of suicidal thoughts in ATX-treated children or adolescents has been described only in one short-term study (Donnelly et al. 2009). A small but significantly increased risk was also observed in a meta-analysis by Bangs et al. (2008b). As in our case, worsening of oppositional behavior could be a marker of distress that could lead to suicidal acts, but no causal link between the emergence of behavioral changes and suicidal impulses has been established (Graham et al. 2011).

During our analysis, two patients, 14 and 8 years old, experienced severe hepatic events. In the first case, the child showed hepatomegaly after 7 months of treatment with ATX. The drug was withdrawn and the clinical conditions improved. Subsequently, the child started a new treatment with MPH, but 2 months later he experienced severe hyperbilirubinemia and hypertransaminasemia. In the second case, hyperbilirubinemia and hypertransaminasemia arose in a child treated with ATX for 10 months. In both cases, the children were not on concomitant medications and, following discontinuation of therapy, liver enzymes levels returned to normal values.

Although the risk of ATX-induced hepatic events has not emerged from clinical trials, some cases of liver injury have been reported during postmarketing experience. Nevertheless, most of these cases contained possible confounding factors or had other explanations unrelated to ATX (Bangs et al. 2008a). Moreover, the probability of under-reporting and the lack of information about the usual prevalence of otherwise unexpected liver injuries, leave genuine uncertainty about the ATX hepatotoxic effect (Graham et al. 2011). The etiopathogenesis of this toxicity is unknown, but an idiosyncratic mechanism was hypothesized (Erdogan et al. 2011).

As with ATX, hepatotoxicity with MPH is a very rare event. A few cases are reported in the literature, but a definite causal relationship has not been established. For the first case described previously, pre-existent moderate liver damage following ATX can be inferred.

Based on the uncertainty about the possible liver toxicity of ADHD medications, our data add important information and may contribute to clarify the safety profile of these drugs.

Compared with in Italy overall, more patients enrolled in Campania dropped out of the ADHD National Register because of at least one ADR (8.4% and 15%, respectively). The most frequent ADRs leading to dropping out were abdominal pain (two serious, one not serious) and weight loss (one serious, two not serious). Among the patients who dropped out, four perceived lack of efficacy. In these four cases, lack of efficacy was not viewed as an ADR and the clinician did not report it to AIFA. According to Rowling and Thompson's ADR Classification (Rawlins and Thompson 1997), updated by Aronson (Aronson 2002), lack of efficacy, as drug failure, is an ADR-subtype and is one of the most frequently under-reported ADRs. Finally, although we did not evaluate treatment efficacy, ∼80% of patients experienced improvement of their clinical conditions, and 24.5% of these discontinued pharmacological treatment for this reason.

The main strength of our study is that, unlike spontaneous reports in the United States Food and Drug Administration's (FDA's) Adverse Event Reporting (AER) system, the authors of this report knew the denominator of the sample and could calculate the AER rate. Our study population represents the entire cohort of Campania children and adolescents with a diagnosis of ADHD, and who take ATX or MPH.

During our study, we introduced monitors who periodically and systematically interviewed clinicians at reference prescription centers. This was an active method to enhance the identification of ADRs by clinicians and to solicit them to report ADRs. The high rate of adverse events in our study population confirms also that an ad hoc form, such as ours, could improve the gathering of data on ADRs, according to Greenhill et al. (2004).

Finally, all ADRs registered in our study population were promptly reported to AIFA, which exceeds the parameters of most of Italian Registries, which generally wait for spontaneous reporting.

Conclusions

Most ADRs from MPH and ATX were those commonly reported in the literature, were not serious, and resolved spontaneously. Nevertheless, several unexpected or uncommon ADRs were identified.

Moreover, our intensive pharmacosurveillance study shows that a close collaboration between the pharmacovigilance centers and clinicians is necessary for improving the effectiveness of the National Register and increasing the number of ADRs recorded in the Italian spontaneous reporting database.

As ADHD is a chronic disorder that typically requires pharamcological treatment for several years or more, further long-term systematic pharmacovigilance studies are needed to identify any rare and unexpected long-term and late-appearing ADRs associated with taking ATX or MPH.

Clinical Significance

Our study investigates a disease whose pharmacological therapy is still being studied for its long-term tolerability. Our active monitoring study allowed us to increase the number of ADRs recorded to the Italian spontaneous reporting database and to collect more thorough information on the safety profile of ADHD medications. Overall, our results provide reassurance of the safety of ATX and MPH in children and adolescents with ADHD. Although our results provide reassurance of the overall safety of ATX and MPH, some unexpected or uncommon ADRs, such as weight gain, hepatotoxicity, suicide ideation, and drug interactions require further investigation.

Disclosures

No competing financial interests exist.