Amfetamine and methylphenidate medications for attention-deficit/hyperactivity disorder: complementary treatment options

The results of the search for published randomized clinical studies that directly compare MPH- and AMF-based stimulants are shown in Table 2. Of the 13 published studies that were identified, there was one parallel-group (i.e. head-to-head) study [76]. This study compared short-acting mixed AMF salts (MAS) with short-acting MPH in the treatment of 58 children with ADHD. As the result of a dose optimization protocol designed to arrive at the ideal balance between efficacy and side-effects, mean daily doses of short-acting MAS and MPH in the final week of the study were 12.5 and 25.2 mg, respectively. The mean daily inattention/overactivity factor of the Inattention/Overactivity with Aggression (IOWA) Conners’ Teacher Rating Scale (CTRS) in the MAS-treated group (mean 0.49) was statistically superior to that of the MPH-treated group (mean 0.81), and both were statistically superior to placebo (mean 1.49). Similarly, the aggression/defiance factor of the IOWA CTRS in the MAS group (mean 0.29) was statistically superior to the MPH group (mean 0.49), and both were statistically superior to placebo (mean 0.72). Further, patients treated with MAS were superior to MPH in the Clinical Global Impressions-Improvements (CGI-I) scale (means 1.6 and 2.35 for the MAS- and MPH-treated patients, respectively) and both were superior to placebo (mean 3.22). The proportion of responders (i.e. improvement in CGI-I scores of 1 or 2) also favoured MAS treatment (MAS 90 %, MPH 65 %, and placebo 27 %) [76]. As will be discussed later, there was no statistical difference in parent-reported side-effects of moderate or severe intensity at the end of the study. Thus, evidence from this parallel-group comparison suggests the superiority of short-acting AMF- over short-acting MPH-based stimulants at optimized daily doses. The authors acknowledged, however, that the dose-optimization algorithm employed, may have limited dosing in the methylphenidate group [76]. It is also possible that the group differences could be a consequence of the longer half-life, and subsequent longer duration of action, of the MAS preparation (up to 6 h) compared to the MPH preparation (up to 4 h).

An advantage of crossover clinical trials is that the within-group design permits the comparison of the treatments in each individual patient, rather than at the group or population level only. A previous comparative review [4] of crossover studies of short-acting formulations of AMF and MPH found no consistent statistical differences in group means of outcome measures. Of a total of 174 patients in six crossover studies, 48 (28 %) responded better to AMF and 27 (16 %) responded better to MPH, and at least 72 (41 %) responded to both; in all studies except for one in which patients exhibited comorbid Tourette’s syndrome, there was a non-significant trend for superior response in patients treated with AMF over those treated with MPH [4].

Most of the crossover studies listed in Table 2 that compared the efficacy of MPH and AMF in patients with ADHD reported equivalence in outcome measures for the two classes of stimulant at the level of the group mean. Those studies that observed statistical superiority of one stimulant over the other for particular outcome measures are reviewed below.

Studies that reported outcomes that favoured MPH included an Australian study, in which 125 treatment-naïve children (aged 5–15 years) were randomly assigned to receive either MPH or AMF [31]. Doses were fixed and based on body weight (short-acting MPH 0.3 mg/kg twice daily; short-acting AMF 0.15 mg/kg twice daily). After 2 weeks of treatment, both stimulants induced significant improvements in baseline scores for all factors of the Conners’ Teacher Rating Scale-Revised (CTRS-R) and the Conners’ Parent Rating Scale-Revised (CPRS-R). In the CTRS-R, there was a statistically significant difference (MPH effect minus AMF effect) in favour of MPH in treatment-induced improvements in conduct problems (difference 3.31, 95 % CI 1.11–5.50, p < 0.01), hyperactivity factor (difference 2.78, 95 % CI 0.70–4.86, p < 0.01), inattentive-passive factor (difference 1.61, 95 % CI 0.30–2.92, p = 0.02) and hyperactivity index (difference 2.60, 95 % CI 0.69–4.51, p < 0.01). In the CPRS-R, the difference in improvement in favour of MPH reached statistical significance for the anxiety factor only (difference 1.20, 95 % CI 0.19–2.20, p = 0.02) [31]. In another study in which short-acting stimulant formulations of MPH 0.45–1.25 mg/kg and AMF 0.2–0.6 mg/kg were administered at breakfast and lunchtime to 18 boys (mean age 9.6 years), both stimulants significantly reduced motor activity (truncal activity counts per hour) compared with placebo, but the reduction was greater for MPH than for AMF between the hours of 11 a.m. and 1 p.m. [12]. A recent crossover study compared the effects of the long-acting stimulant formulations osmotic release oral system MPH (OROS MPH; maximum daily dose 72 mg) and extended-release MAS (maximum daily dose 30 mg) on neuropsychological functioning in adolescents with ADHD (n = 35; mean age 17.5 years) [98]. There were no significant differences between OROS MPH and extended-release MAS in any of the outcomes studied. However, OROS MPH, but not extended-release MAS, was statistically superior to placebo in distracter errors and distracter reaction time in the Go/No-Go test and in Recall Accuracy in the Delayed Matching-to-Sample test, perhaps reflecting greater than a twofold MPH to AMF dose-ratio than that considered to be equivalent [98, 101].

In contrast to the above results favouring MPH, responses to short-acting AMF (maximum daily dose 45 mg) were superior to short-acting MPH (maximum daily dose 90 mg) in a crossover study of classroom performance in boys aged 6–12 years with ADHD. Compared with placebo, both drugs produced statistically significant improvements in performance (percent correct responses) and number of problems attempted for reading tasks, and both drugs improved the number of attempted arithmetic problems. However, improved performance in arithmetic problems compared with placebo was statistically significant for AMF only (mean [standard deviation] percent correct: AMF 97.1 [4.6]; MPH 96.2 [5.6]; placebo 94.0 [7.9]) [36]. Again, the longer duration action of the amphetamine may have influenced these results.

Using standardized effect sizes, it is possible indirectly to compare efficacy outcomes for particular treatments across studies. Standardized mean difference (SMD) is one way of calculating effect size. For example, an SMD of 1 indicates that the mean outcomes in drug and placebo groups differ by 1 (pooled) standard deviation. In interpreting SMD values, an SMD of 0.2 is considered small, 0.5 is considered medium, and ≥0.8 is considered large [21] (Fig. 2).

A series of recent meta-analyses have examined effect sizes of various outcome measures in patients with ADHD of different ages treated with different ADHD medications [38‐40]. In a meta-analysis of 32 double-blind, placebo-controlled trials in youths (aged 6–18 years) with ADHD, meta-analysis regression found that effect sizes for non-stimulant medications (SMD = 0.57) were significantly smaller than for both short-acting (SMD = 0.99, F_1,31 = 25, p < 0.0001) and long-acting stimulants (SMD = 0.95, F_1,31 = 15, p < 0.0001) [38]. Although this analysis did not stratify stimulant therapies according to their MPH or AMF class, it is interesting to note that the largest effect sizes for both short-acting (MAS, SMD = 1.34) and long-acting (LDX, SMD = 1.52) stimulants were seen for AMF-based therapies. When effect sizes were compared across 23 randomized, controlled trials of 11 different short- and long-acting stimulants in children and adolescents with ADHD, robust drug effects were observed in all individual studies. Furthermore, across all of these studies, effect sizes for AMF products were significantly greater than for MPH products (SMD = 1.03 vs. 0.77, t ₁₉ = 2.5, p = 0.02). Regression analyses found that several study variables, including stimulant formulation (short-acting vs. long-acting), type of dosing (fixed vs. optimized), and study design (parallel vs. crossover), were not associated with SMD. It is, of course, possible that heterogeneity between study variables may have obscured possible associations between these variables and SMD. Nevertheless, three study design features were identified that they were associated with SMD (age, type of score [outcome or change], and rater [physician, parent, teacher, or patient]), but the finding that effect sizes for AMF were modestly greater than those for MPH held after correcting for these confounding variables [39]. Equivalence of dosing between AMF and MPH stimulants was not demonstrated. Using the method of numbers needed to treat, another way of comparing outcomes from different studies, the authors calculated that 2.0 patients were needed to be treated with AMF for each positive outcome for total ADHD symptoms compared with 2.6 patients for MPH [39]. Overall, these pooled analyses provide strong evidence for the efficacy of stimulant medications in ADHD, while one meta-analysis provides evidence of the greater efficacy of AMF-based drugs compared with MPH-based drugs in children and adolescents. Again, the longer duration action of short-acting AMF than short-acting MPH may have contributed to these results.

Overall, some individual studies have demonstrated superiority of MPH over AMF, some have found superiority of AMF over MPH, and others have shown no differences between the two types of medications. When meta-analyses are performed summarizing all available evidence, the effect sizes observed with AMF are greater than those observed with MPH, although issues of comparable dosing and differences in the duration of action of short-acting stimulants should be taken into account when interpreting these data. However, given the currently available evidence, it has not been demonstrated that one stimulant is more efficacious than another at a population level; direct head-to-head studies would be required to establish this definitively.

Although group average responses to MPH and AMF in patients with ADHD are similar, individuals may respond very differently to the two drugs. While approximately two-thirds of patients typically experience improvements in various symptom domains in response to a single stimulant, if patients with an unsatisfactory response try the alternative class of stimulant, the proportion of patients who respond to one of the drugs may be as high as 95 % [4, 31]. Updated response data for studies directly comparing MPH- and AMF-based stimulants are presented in Table 2. Of eight studies containing data for MPH and AMF formulations, the proportion of responders was higher for AMF in four studies, higher for MPH in three studies, and were equivalent in one study. In the only head-to-head comparison of stimulants, 90 % of patients responded to MAS, 65 % responded to MPH, and 27 % responded to placebo [76]. When numbers of responders for each stimulant were combined across studies, 226 of 318 patients (71 %) responded to MPH and 216 (68 %) responded to AMF, suggesting that there is no meaningful difference in numbers of responders for MPH and AMF in ADHD. However, the proportion of patients responding to either class of stimulant (287 of 316 patients 91 %) was higher than those responding to each single stimulant.

These analyses confirm previous assertions that non-response is uncommon when an individual is offered both a MPH and an AMF [35], and that responses to the two classes of stimulant, although similar in the overall ADHD population, does vary between individuals [4]. Differences in the metabolic pathways and mechanisms of action of MPH and AMF (see above), the genotype of an individual [44, 89] and the pathophysiology of their ADHD [26] may be important factors in determining an individual’s response to the different stimulant drugs. In terms of clinical practice, these differential response rates support clinical guidelines that recommend that MPH and AMF are equally valid first-choice medications for the treatment of ADHD, and that if the first-selected stimulant class proves to be unsatisfactory, then a stimulant from the second class should be tried [75, 77].

The selection of which stimulant class to start with may be aided by methods for identifying patient subgroups that may preferentially respond to medication, including pre-specified and post-hoc subgroup analyses of clinical data [11, 23, 28, 29, 54, 87]. However, there are well-accepted constraints of both a priori and post-hoc specification [8, 56, 74, 78]. These limitations are starting to be addressed by the development of personalized treatment selection techniques; these combine patient characteristics from clinical trial data to form a risk score and then performing analyses on subgroups using a two-stage process that allows for treatment responses at an individual level to be made [16, 94, 102]. These methodologies have the advantage that they can systematically use multivariate regression models to select and combine multiple baseline covariates from different levels of analysis to define subgroups. With a view to the future, pharmacogenomics and non-genetic biomarkers may assist in the optimization and individualization of ADHD pharmacotherapy, although this is not yet possible [10, 19, 34, 44, 50, 89].

Reductions in the symptoms of ADHD by stimulants depend upon achieving sufficient occupancy of their molecular target(s) in the brain [95]. Positron emission studies suggest that peak occupancy of the dopamine transporter is reached approximately 60 min after oral dosing with short-acting MPH [96]. However, the elimination half-life of short-acting MPH is reported to be approximately 3 h [51] and of AMF to be approximately 7 h [14]. Therefore, two or, in the case of short-acting MPH, three daily doses are necessary to maintain therapeutic concentrations of stimulants within the brain throughout the day. For children and adolescents, repeat dosing may be undesirable because of the difficulties associated with storing and administering scheduled drugs within a school environment, the stigma associated with receiving medication during the school day, fragmented coverage with multiple short-acting doses, the potential for diversion of drug and the possible impact on adherence to the dosing regimen [100].

To extend the duration of action (i.e. symptomatic control) of stimulant medications, several long-acting formulations of MPH and AMF have been introduced that extend the pharmacokinetic and pharmacodynamic profile of the drugs (Table 1) [43, 61]. Most formulations depend on the slow, sustained release of the active ingredient in the stomach via the use of technologies such as wax matrix tablets, capsulated biphasic beads, or osmotically controlled release systems. Long-acting formulations mean that systemic exposure, and hence efficacy (symptom control), is maintained for longer periods, resulting in correspondingly improved convenience, confidentiality and compliance, more consistent coverage, and reduced abuse potential [43, 61, 75]. LDX is the first stimulant to use prodrug technology to modify the delivery profile. In its parent form, LDX is inactive and requires enzymatic cleavage in the blood to yield AMF. The combination of short- and long-acting formulations provides a range of treatment options lasting from approximately 4 h to more than 12 h. The demonstration that the efficacy of LDX in children is maintained for at least 13 h [97], suggests that this prodrug is the longest-acting stimulant formulation [25, 45]. However, since the drug needs to be absorbed and then cleaved in the bloodstream before it can be active, the onset of action may be somewhat delayed and may occur 1.5–2 h after ingestion [97]. In selecting an ADHD medication, stimulants may be contraindicated or the patient or caregivers may express a preference for non-stimulants. In such cases, a non-stimulant such as atomoxetine may be considered. Generally, the non-stimulants are considered less effective than the stimulants. Where stimulants are considered appropriate, the choice of short- or long-acting stimulant should be based on both clinical requirements and the preferences of an individual and their family [9].

Despite the carefully managed nature of the Multimodal Treatment Study of Children with ADHD (MTA), saliva assays for MPH revealed that approximately 25 % of 254 patients in the medication arms of the study were non-adherent on 50 % or more of repeated assays, and that barely half (54 %) were adherent at every assay point [67]. Furthermore, discrepancies were uncovered between parents’ reports of adherence and the outcomes of the saliva assays [67]. These results indicate that there is considerable potential to improve pharmacotherapy outcomes by improving adherence. Several strands of evidence suggest that medication adherence may be improved by tailoring the selection of drug to an individual patient. Discrete choice experiments suggest that long-acting stimulants, with consistent therapeutic coverage throughout the day, are the preferred stimulant formulations of most patients [20, 42, 58, 64], and retrospective claims analyses suggest that long-acting stimulants are generally associated with enhanced adherence and persistence in patients of all ages compared with short-acting stimulants [20, 59, 80, 83]. In addition to drug regimens that patients find convenient, other strategies to improve adherence include improved communication between physicians, caregivers, and patients; clear instruction and encouragement; peer support groups; advice about reminders to take medication and incorporating medication into daily routines; and the use of positive reinforcement to improve attitudes to medication [66]. Addressing adverse events effectively may also contribute to promoting adherence.

The importance of utilizing a multimodal treatment strategy that incorporates both medication and non-drug interventions is recognized by all ADHD clinical guidelines [1, 17, 66, 75, 91]. Indeed, as mentioned previously, non-drug interventions are the first-line treatment in school-aged children and adolescents with moderate ADHD and moderate impairment in many European countries [66, 91]. Behavioural therapy alone can produce improvements in ADHD compared with baseline [73]. In the MTA, for example, both medication and intensive behavioural therapy provided superior treatment outcomes to treatment in the community, even though the community treatment often included medication [92]. As the medication arm of the MTA was superior to both the behavioural and community treatment arms, these data provide further support for the notion that for medication treatments to achieve optimal effectiveness they need to be carefully titrated and monitored. The MTA also demonstrated that there were benefits in combining medication and behavioural therapy in certain non-core ADHD symptom domains (including aggression, internalizing symptoms, social skills, and parent–child relations), compared with either treatment approach alone [92]. Furthermore, non-adherence resulted in greater deleterious effects in the medication management group than in the combined treatment group [67]. The results from the MTA study confirmed those of earlier studies in children attending summer treatment programmes that demonstrated the reinforcing and synergistic outcomes of pharmacotherapy and behavioural therapy [18, 70, 71]. Together, these data illustrate the potential therapeutic advantages of combining intensive behavioural therapy with carefully crafted medication management [81].