Ambermoon and colleagues provide a brief overview of the epidemiology and nature of dopamine dsyregulation syndrome (DDS), which occurs in a small percentage of Parkinson's disease (PD) patients on dopamine replacement therapy (DRT) [1]. Ambermoon and colleagues reiterate that study of these patients could inform the field regarding the pathogenesis and neurobiology of addiction more generally. They also suggest that the compulsive use of DRT by PD patients provides a useful model of stimulant addiction, although the depth at which this proposal is examined is somewhat lacking.

What is apparent is that a lack of clear classification and standardized diagnostic criteria for reporting is impeding progress in both the study and treatment of DDS, and compulsive use of DRT in particular. Classifying compulsive DRT use as a ‘Dopaminergic Medication Use Disorder’ (DMUS) would bring it into line with the proposed DSM-V classifications [2], allowing the formalization of more addiction-specific diagnostic criteria. This would assist in reframing of the disorder as an ‘addiction’ which would then have flow-on effects in treatment and research. Nevertheless, until compulsive DRT use is distinguished from other features of DDS it is difficult to assess its validity as a model of stimulant addiction. More detailed reporting will assist the field in determining the exact nature of each aspect of DDS and their inter-relationships, allowing a more robust appraisal of potential common pathogenesis. The fact that compulsive DRT use is associated more with l-dopa, and impulse control disorders (ICDs) more with dopamine agonist therapy suggests that there are likely to be differences which warrant investigation.

Determining the relationship between compulsive DRT use and other features of DDS, such as ICDs, is important. For example, are ICDs more common when there is escalation of DRT use? To what extent are they comorbid? Does one precede the other? Are the ICDs essentially a result of the DRT medication sensitizing reward circuitry, impairing reward learning and/or a result of prefrontal deficits in inhibitory control? Assessing general impulse control in these patients on tasks such as Go/NoGo or Stroop tasks would help; yet the cognitive deficits which occur in PD are an obvious confound. Given that the degeneration of a major striatal input pathway in PD, as well as the existence of cognitive deficits, how useful is the comparison to psychostimulant addicts without PD? An alternative option would be the investigation of compulsive DRT use in disorders with less potential confounds such as restless leg syndrome. In addition, regardless of the patient group in question, the pattern of drug administration is quite different with DRT versus psychostimulant use, with the former being stable and relatively continuous and the latter typically being episodic and binge-like. These differences also make comparisons between the two groups potentially problematic.

Regardless, the examination DDS in PD patients has yielded invaluable information regarding the neurobiology of addiction and there is ongoing research in this area (for reviews see [3–5]). First, the observation that only a small number of patients are affected indicates the presence of pre-existing vulnerabilities which predispose certain individuals to these pathological behaviours. It appears that, like drug addicts, PD patients with predispositions to addictive behaviour, are more sensation- and novelty-seeking [6] and have lower levels of D2 receptors in the striatum [7]. They have exhibited sensitized responses to l-dopa in the ventral striatum [8] and exhibit heightened responses to reward-related cues [9]. Work with PD patients supports elegant experiments which have informed our knowledge on the role of dopamine in reward learning and striatal plasticity [4]. Thus, it appears that the field has, for quite some time, recognized the utility of studying PD patients in this context and that Ambermoon and colleagues are emphasizing the need to separate compulsive DRT use from other features of DDS.

There remain key experiments which will help us advance knowledge in this field. Investigation into whether or not deficits in cortico-striatal plasticity exist in this model will help to determine whether or not ‘addiction’ in this context is akin to psychostimulant addiction. In addition, as Ambermoon and colleagues suggest, imaging patients before the initiation of treatment will provide valuable information that is impossible to obtain from drug users, potentially providing researchers with a useful biomarker.

Declarations of interest