Response

Psychopharmacological treatments influence neuronal function and mental states through effects on the cells of the central nervous system. Many compounds with potentially therapeutic effects on psychiatric or neurological conditions are excluded by the blood–brain barrier. In their article, Davies and colleagues consider the potential of one such compound – the potassium-sparing diuretic amiloride for the treatment of panic disorder. ^{Reference Davies, Schruers and Liebold1} They consider the potential of an intranasal formulation to overcome the poor permeability of amiloride across the blood–brain barrier.

We require new pharmacological treatments in psychiatry with greater efficacy and faster onset of action. This is particularly the case in anxiety disorders, where the translation of anxiolytic compounds from animal models of anxiety into the clinic has been very limited over recent decades. ^{Reference Griebel and Holmes2} For panic disorder, the last major addition to the available pharmacological treatment was the introduction of the serotonin-noradrenaline reuptake inhibitors in the 1990s and 2000s.

The rationale presented for amiloride in potentially ameliorating panic symptoms appears strong. The blockade of ion channels in the kidney underlies its diuretic effect, while the additional antagonism of acid-sensitive ion channels (ASICs) has the potential for an anxiolytic action. Genetic variants in one of the genes encoding ASICs (the human ACCN2 gene), and alterations in gene expression in peripheral tissue, are associated with increased reactivity to carbon dioxide inhalation, a long-standing experimental model of panic. Inhalation of air enriched with carbon dioxide results in the physiological and psychological symptoms of panic, and the physiological features are replicated in rodent studies. In preclinical studies in mice exposed to a model of early life adversity, pre-treatment with nebulised amiloride reduced the physiological response to carbon dioxide inhalation.

Amiloride has been used as a diuretic for many decades, with no anxiolytic effect being recognised, in line with its predicted poor permeability of the blood–brain barrier. This is unfortunate considering the high rates of anxiety disorders and clinically significant levels of anxiety symptoms in heart failure. ^{Reference Celano, Villegas, Albanese, Gaggin and Huffman3} Intraperitoneal delivery of amiloride in the mouse study described above had no effect on the response to carbon dioxide inhalation. This is in keeping with a human study in 51 healthy volunteers exposed to a 20 min inhalation of 7.5% carbon dioxide: ^{Reference Parker, Mahood, Taylor, Mills, Hou and Baldwin4} pre-treatment with oral amiloride (10 mg) showed no significant effect on anxiety symptoms versus placebo. However, the 7.5% carbon dioxide challenge is considered an experimental medicine model of generalised anxiety disorder rather than panic disorder per se. ^{Reference Bailey, Dawson, Dourish and Nutt5} Overall, targeted delivery to the central nervous system appears essential for the potential efficacy of amiloride.

The intranasal route of administration could potentially provide this route for compounds with poor permeability to the blood–brain barrier. Intranasal administration relies on direct transportation of compounds to the brain via olfactory and trigeminal nerve fibres, through both endocytosis into the neurons and extracellular diffusion. ^{Reference Quintana, Guastella, Westlye and Andreassen6} Additionally, rapid uptake by capillaries in the mucous membranes of the nose occurs. From there, compounds enter the systematic circulation and can cross the blood–brain barrier, depending on their permeability. There has been increasing interest in the development of psychotropic formulations delivered in this way in recent years – for example, the introduction of intranasal esketamine in treatment-resistant depression ^{Reference McIntyre, Rosenblat, Nemeroff, Sanacora, Murrough and Berk7} and intranasal naloxone for acute treatment of opioid overdose. Intranasal formulations of the benzodiazepines midazolam and diazepam have also been developed for administration during epileptic seizures. It is worth noting that these compounds have significant blood–brain barrier permeability, and the intranasal delivery system is of benefit by allowing rapid uptake by the systematic circulation, thereby avoiding first-pass metabolism in the liver. The neuropeptide oxytocin, however, is an example of a compound with apparent poor permeability across the blood–brain barrier. Intranasal administration has been widely studied in humans and animals, where evidence appears to support direct transportation to the brain via olfactory and trigeminal nerve fibres. ^{Reference Quintana, Lischke, Grace, Scheele, Ma and Becker8}

Amiloride nasal spray, discussed by the authors, has been studied in one healthy volunteer study in humans to date. It enters the systematic circulation rapidly, reaching peak plasma concentrations after 10 min, at the latest, in most participants. At this stage, without the direct measurement of cerebrospinal fluid concentrations (as done for example in intranasal oxytocin administration ^{Reference Quintana, Lischke, Grace, Scheele, Ma and Becker8} ) or relevant psychological or physiological measures of panic symptoms, it is difficult to confirm the delivery of amiloride to the human central nervous system as predicted by theoretical models. Intranasal administration in humans also differs in some ways from the manner of delivery in animal studies, where mice were placed in a chamber containing nebulised amiloride.

Results from further human studies using intranasally administrated amiloride are eagerly awaited. Studies in both experimental anxiety models with healthy volunteers (such as carbon dioxide inhalation) and patient populations with panic disorder could be conducted. Theoretically, polyuria and hyperkalaemia could occur due to the renal effects of amiloride, and this would need to be monitored, particularly with repeated dosing. If the potential of amiloride is realised, it will further support the option to deliver compounds with poor blood–brain barrier permeability into the central nervous system. This could open the way for other such compounds with established licences outside psychiatry or neurology to potentially be repurposed for conditions in these fields. Finally, and specifically for panic disorder, a rapidly acting anti-panic compound could markedly reduce the associated distress and disability of panic attacks and improve the lives of those affected by the disorder.