Refining Treatment Choices for ADHD

As diagnostic rates of attention-deficit hyperactivity disorder (ADHD) increase worldwide, the most common questions asked by patients are whether to start a medication, which one, and for how long? In the Lancet Psychiatry, Samuel Cortese and colleagues report findings of a comprehensive network meta-analysis¹ in which they synthesized crucial information for development of guidelines that will inform clinical decision-making. By including recent studies published up to April 2017, and previously unpublished information, Cortese and colleagues’ results clarify inconsistencies in earlier reviews and meta-analyses, some of which have generated much controversy.^2,3 Furthermore, Cortese and colleagues assessed secondary safety and tolerability outcomes, reflecting growing recognition that acute symptom reduction is only half the battle in treatment of ADHD, which ultimately should target impairment or function^4,5.

For highly prevalent disorders such as ADHD, guidelines for guidelines for clinical care should be informed by both public health and precision medicine approaches⁶. The network meta-analysis by Cortese and colleagues directly addresses the public health approach to pharmacologiclal intervention and confirms the robust short-term efficacy of stimulant medications, although not all medications had the same effects. Moreover, their results present useful clinical indications for close monitoring of blood pressure and weight changes, not only with stimulants but also with atomoxetine.

From a precision medicine angle, the findings of the network meta-analysis were limited by the paucity of existing longitudinal and active comparator studies. The results support a moderating effect of age on first-line medication choice, consistent with current use patterns in the USA⁷, with greatest support for methylphenidate in children and adolescents and amphetamine formulations in adults. However, the analyses lacked power to further differentiate effects among children (ages 5–12) versus adolescents (ages 13–18). Thus, the network met-analysis was unable to address a major challenge in ADHD treatment, which is the high rate of discontinuation that occurs around puberty despite continuation and even increased severity of impairments at that age.⁸

The network plots in the current network meta-analysis, in which placebo dominates the picture, underscore the need for a paradigm shift toward comparative treatment research. Despite evidence for preferential response to either methylphenidate or amphetamine at the individual level⁹, we still cannot predict a patient’s response to first- or second-line treatments, let alone indications for multiple interventions in cases of partial response. In terms of treatment algorithm development for a frequently chronic problem for which monotherapy is seldom sufficient, we remain stuck at the very first branch.

This network meta-analysis presents an advanced methodology to summarize and compare findings from treatment studies. Regarding the three common questions patients ask, the extant literature meant that the network meta-analysis could derive only partial answers, including general support for short-term medication use, basic guidelines about first-line medications based on age, and little to no data for duration of treatment. These results – and particularly limitations of the data obtained – call for future research that translates directly to clinical practice. Ultimately, this need will require funding agencies to support longitudinal and active comparison designs, attention to sequencing and combining pharmacological and behavioral interventions¹⁰, and development of big datasets that permit identification of patient predictors, which have thus far remained elusive.