Restless legs syndrome, neuroleptic-induced akathisia, and the iron opioid dopamine link

We read with interest the work of Ferre et al. [1]. The authors point out the clinical similarities of restless legs syndrome (RLS), neuroleptic-induced akathisia (NIA), and opioid-withdrawal restlessness. All three conditions are characterized by restlessness and therefore may share a common pathophysiology. In addition, they note the similarity in response of all three conditions to opiate therapy and, based upon rodent experiments, they suggest that specific striatal neurons that express mu opiate receptors are involved in the pathogenesis of all three conditions [1]. The authors implicate gamma-aminobutyric acid, glutamate, adenosine, dopamine, and endogenous opiates as being important in pathogenesis of restlessness [1]. Other authors have also implicated noradrenaline and serotonin in NIA [2] and in RLS [3^, 4] Iron is also low in RLS as has been proven by multiple methods of experimentation including serum, cereobrospinal fluid (CSF), MRI, and brain autopsy studies [5]. A low iron status is also implicated in the pathogenesis of NIA. In a meta-analysis of 10 studies in human participants with NIA, low iron status was mostly related to NIA [6] but negative studies have been reported as well [7^, 8].

Ferre et al. have laid down some convincing facts linking the endogenous opiate system to the pathogenesis of restlessness of all types—RLS, NIA, and opiate withdrawal. The authors have also laid out some reasonable and perfectly testable hypotheses for future research [1].

A polysomnographic study compared the clinical differences between RLS and NIA [9]. Both conditions are characterized by increased motor activity. In RLS, patients may walk on the floor, stretch or bend their legs, toss and turn in bed, or provide counterstimulation such as rubbing the legs in response to leg discomfort. In NIA the restlessness is driven more often by an inner restlessness rather than leg discomfort. In patients with RLS the strategy of which ways to move to respond to leg discomfort are chosen at will but in NIA the movements are often more stereotypic such as marching in place and body rocking. In RLS symptoms are more prominent at night but in NIA they may be present equally day and night [9^, 10]. In RLS, symptoms are more prominent on sitting or lying but this is not necessarily true in NIA [9^, 10]. Patients with RLS and NIA may both have sleep disturbances and periodic limb movements of sleep but periodic limb movements of sleep seem to be more characteristic of RLS than NIA [9^, 10].

To make things more complicated, however, NIA comes in three forms. There is an acute form (acute NIA) which is generally reversible by a variety of therapeutic agents and by the discontinuation of dopamine receptor-blocking agents. There is also a tardive form (tardive neuroleptic akathisia) where the restlessness of NIA persists and may even become worse despite the discontinuation of neuroleptics. There is also a third form, pseudoakathisia, where the restless movements persist but the feeling of restlessness has disappeared [11^, 12]. Acute NIA usually occurs after brief periods of therapy whereas tardive and pseudoakathisia are usually consequences of long-term therapy with neuroleptics [11^, 12]. In tardive NIA, there is also a generalized resistance to therapy of all types, and this may be due to permanent changes in dopamine receptor sensitivity that occur with long-term exposure to neuroleptics [13]. Future research should take into account the phenotypic differences between these three types of NIA.

Ferre et al. extend our early observations by exploring the role of the endogenous opiate system in the production of restlessness. It has been well-proven that opioids ameliorate RLS [14^, 15]. We showed that two opioid-treated patients with RLS lost the benefit of opioids upon their RLS symptoms when naloxone was administered i.v. in a double-blind fashion as compared to saline [16^, 17]. It must be stated; however, that naltrexone has no effect on RLS symptoms when administered in the absence of opioid therapy [18]. Whether this relates to the differential strength of binding of endogenous as opposed to exogenous opioids to the opioid receptor or other factors is at this time unknown. Other evidence supporting the role of the endogenous opiate system in the pathogenesis of RLS is elaborated in a recent publication [5].

We then followed up these studies with similar studies in patients with NIA. As with RLS, opioids improved the symptoms of restlessness in NIA in a double-blind study. Furthermore, the improvement of one of the participants while on opioids was reversed when naloxone versus saline as a placebo were administered in a double-blind fashion [19]. The videotape showing the remarkable reversion to symptomatology with naltrexone while on opioids and the lack of reversion with saline while on opioids is preserved in the archives of the Movement Disorder Society [19]. We later published a follow-up analysis to our original study [13^, 19]. The original study contained a mixture of acute and tardive narcoleptic patients [19]. In the follow-up analysis, we noted that although acute NIA and tardive NIA both respond to opioid therapy, responsiveness was less in the tardive group [13]. Again, this may relate to permanent changes in receptor sensitivity after long exposure to neuroleptics.

There is one criticism that can be offered for the hypotheses put forward by Ferre et al. [1]. In their comparison of RLS, NIA, and opiate withdrawal, the authors rely heavily on literature that suggests synaptic dopamine levels are high in RLS. In autopsy studies, patients with RLS show a decrease in D2 receptors in the Putamen and a substantia nigral increase in tyrosine hydroxylase, the rate-limiting step in dopamine synthesis [20]. The authors of these studies suggest that this implies that dopamine levels are actually high in RLS. Some but not all neuroimaging studies are also supportive of the hyperdopaminergic hypothesis. However, other interpretations of this data are possible [5]

• (1) Firstly, it is important to note that the authors of the autopsy studies did not actually measure dopamine levels.

• (2) Secondly, the increase in tyrosine hydroxylase may be compensatory for low dopamine levels.

• (3) Thirdly, it is possible that the decrease in dopamine receptors may be intrinsic to RLS and not just a change secondary to increased dopamine levels.

• (4) Fourthly, in RLS dopamine and its metabolites are highest in CSF in the morning when RLS symptoms are at their minimum. Thus it could be that dopamine levels are high overall but a decrease in dopamine levels at night could make patients more susceptible to RLS symptoms at that time.

The hyperdopaminergic hypothesis for restlessness also runs directly counter to other observations:

• 1 The first is that dopamine agonists increase functional dopaminergic activity and they are the ones that lower the restless activity in RLS and not the dopamine antagonists which decrease functional dopaminergic activity. This again is compatible with a low (not high) dopaminergic state in RLS [5].

• 2 At least in acute NIA it is the dopamine receptor-blocking antipsychotics and dopamine receptor-blocking GI medications such as metoclopramide that cause the restless symptoms of NIA by causing a low (not high) dopamine state [2^, 6–9^, 11–13^, 19].

• 3 In in vitro animal models opioids rescue dopaminergic cells from a low (not high) dopaminergic state that is induced by iron deficiency [21].