Can D2-based therapies fix presynaptic dopamine?

The pursuit of better treatment for schizophrenia has motivated a large body of work in biological psychiatry. One of the most well established findings is the presence of excess dopaminergic function in the striatum (1), and has led to ongoing discussion in the field of whether psychosis may be fundamentally a pre- or post-synaptic dopamine (DA) dysfunction (2). Current antipsychotic treatments block dopamine D2 receptors (D2), but it is possible that the therapeutic mechanism of D2 blockers is to reduce presynaptic DA. This is particularly important since D2 blockers, when administered chronically, seem to induce D2 hypersensitivity, necessitating further D2 blockade for effective treatment of chronic psychotic disorders. However, if the primary abnormality in psychosis is excess presynaptic DA, then drugs that directly reduce DA could prove to be effective. Testing this is complicated by the pathophysiology of the illness itself, which may make therapeutic intervention a moving target. Molecular imaging of first episode psychosis may help us to address these important questions and has the potential to identify specific neurochemical targets for novel treatments of psychosis before it develops into a chronic illness.

Excess presynaptic DAergic function in schizophrenia has been observed using various imaging approaches, one of which is thought to reflect dopamine synthesis capacity (DSC) (1). Several prior studies have observed excess striatal DSC in patients with schizophrenia, and the associative and limbic striatal subregions appear to have the greatest excess in DSC in antipsychotic-responsive patients with schizophrenia (1). More recently, excess DSC has been identified in patients with psychosis, irrespective of diagnosis of bipolar disorder or schizophrenia (3). In a report from McGowan et al (4) of excess DSC in a group of patients with schizophrenia compared to HCs, the patients were all on antipsychotic medications and minimally symptomatic, suggesting that DSC may not be affected by antipsychotic medication --but the patients’ pre-treatment illness severity could not be accounted for in the cross-sectional study design, so it is difficult to rule-out the possibility that excess DSC may have been reduced by antipsychotic medication.

This issue of Biological Psychiatry includes the first longitudinal study examining DSC before and after the treatment of first-episode psychosis. In Jauhar et al. (5), 17 patients suffering from their first psychotic episode underwent positron emission tomography (PET) imaging with 3,4-dihydroxy-6-[¹⁸F]fluoro-L-phenylalanine ([¹⁸F]DOPA) radioligand before and after several weeks of treatment with second-generation antipsychotic medication.

The authors hypothesized that DSC—a proxy measure of presynaptic DAergic function—would be initially elevated and decrease with effective antipsychotic treatment, as might be expected given multiple observations of the correlation of the severity of psychosis in patients with schizophrenia with excess DSC (1), suggesting that this measure might reflect a state-component of psychosis. Thus, if psychosis decreases, DSC might decrease as well. Indeed, the only prior longitudinal study of DSC with antipsychotic treatment by Grunder et al. (6) showed that striatal DSC in patients with schizophrenia decreased by 25% after effective treatment. Furthermore, excess DSC has been observed in treatment-responsive, but not in treatment-resistant patients with schizophrenia, suggesting that a treatment responsive subtype of the illness is associated with abnormalities of DSC that correct with antipsychotic medications.

Preclinical studies make the story more interesting. In wild-type rodents, when administered chronically, antipsychotic medications— all D2 blockers—silence the firing of a number of midbrain DA neurons via the prolonged antipsychotic binding to D2-like autoreceptors. This result is dependent on striatal input (see (7)), implicating a mechanism whereby antipsychotics reduce presynaptic DAergic function, and, in turn, DSC. In contrast with expectation, however, Jauhar et al. report that DSC remains high in patients with first episode psychosis, even in the presence of a significant therapeutic response to antipsychotic medication: after 5 weeks of antipsychotic treatment, no changes were observed in striatal DSC (indexed as [¹⁸F]DOPA influx rate constant, K_I^ceri). Nor were there observable treatment-related changes in K_I^ceri in the thalamus or midbrain.

Replicating previous findings, Jauhar et al. did see a significant positive relationship between the baseline imaging measure K_I^cer_i in associative striatum and positive symptom severity change with treatment. However—and contrary to what one might expect if excess DSC was a state measure of psychosis—the decrease in positive symptom severity in individual subjects was not correlated with changes in K_I^cer_i. This result conflicts with that of Grunder et al., whose study sample, it must be said, was significantly less symptomatic, included patients with schizophrenia but none with psychosis secondary to bipolar disorder, and were all treated with first generation antipsychotic with high D2 occupancy. Moreover, and perhaps most importantly, Grunder et al. largely studied chronic patients; this is likely a germane difference given that sensitization models of schizophrenia and its treatment (8) suggest that the dopaminergic system is affected by chronic psychosis and by exposure to D2 blockade, such that individuals in their first episode of psychosis may respond differently to antipsychotic medications.

If we interpret Jauhar et al.’s findings as suggestion that DSC is not affected by chronic D2-blockade or reduction in psychotic symptoms, we might infer that excess presynapticDAergic function is part of the illness and that the therapeutic mechanism of antipsychotic medication is to reduce postsynaptic transmission by blocking D2. This provides rationale for continuing antipsychotic medications, even after symptoms resolve, to maintain D2-blockade and prevent relapse. However, we still do not know that chronic treatment does not further alter the DAergic system. For example, if chronic antipsychotic treatment results in D2 hypersensitization, it is possible that DSC may increase concordantly and thereby indicate increased risk for symptom relapse. It will be important to follow the course of illness for these individuals and repeat DSC measurements longitudinally to determine whether DSC tracks with other aspects of the course of illness or its treatment.

Also, given that there are multiple biochemical contributions to the DSC imaging measure, we do not know whether antipsychotic-induced decreases in phasic DA release may be offset by increases in tonic DA activity, and so we cannot rule out the possibility that the various contributions to the DSC imaging measure are not rebalanced in some way, while still maintaining a net zero change of overall DSC.

Since no reduction in excess presynaptic DAergic function was observed, this study cannot tell us whether lowering presynaptic DA may reduce psychosis. It is still possible that by lowering presynaptic DAergic function (e.g. DA synthesis or release), psychotic symptoms may be treatable without blocking D2 postsynaptically. To determine whether psychotic disorder may be fundamentally a pre- or post-synaptic DA dysfunction, we need experimental paradigms that can distinguish between pre- and post-synaptic DAergic (dys)functions in patients. Proof of concept studies such as DA depletion versus D2 blocking strategies will be crucial. If modulating presynaptic DA can reduce severity or progression of psychotic illness, it would be exciting to find out whether pharmaceuticals that modulate presynaptic DA, such as Trace Amino Acid Receptor 1a (TAAR1a) (9) or Kappa-opioid receptor (10) –both of which are in the pipeline for use in humans—can be an effective treatment.