Why ASIC antagonism might suppress panic

Preclinical, clinical and pharmacological data support a pathophysiological role for ASICs, and hence a therapeutic role for the ASIC antagonist amiloride.

People with panic disorder undergo measurable respiratory and autonomic instability such as changes in tidal volume and CO₂ partial pressure, which often begin up to 50 min before the consciously detectable onset of a panic attack. ^{Reference Meuret, Rosenfield, Wilhelm, Zhou, Conrad and Ritz14} Although these respiratory fluctuations (most prominently hyperventilation) are driven by extracellular pH changes, they probably reflect hypersensitive pH-detecting functions that feed into the generation of full-blown panic attacks. ^{Reference Meuret, Rosenfield, Wilhelm, Zhou, Conrad and Ritz14–Reference Leibold, van den Hove, Esquivel, De Cort, Goossens and Strackx16} Conscious awareness of bodily signals develops only later in the genesis of a panic attack, after which the subjective experience of emotional and cognitive symptoms may begin. Once this point is reached, a targeted, fast-acting therapeutic intervention could limit panic attacks before they reach the peak of symptom intensity ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12} within the timespan of ‘minutes’ that DSM-5 defines as the maximum interval between awareness of panic attack onset and its peak. ⁵

The possibility that panic attacks might be ‘intercepted’ before they can cause maximal clinically significant distress is plausible once one identifies (a) a pathophysiological mechanistic target, (b) a molecule that acts on this target and (c) an administration route allowing this molecule to reach the brain quickly and exert its effect once the initial panic attack signs are evident. The blood-acidifying signals that occur as panic attacks develop can be reproduced in the laboratory using paradigms such as CO₂ stimulation via respiratory challenges (inducing hypercapnia, and hence relative acidosis in the extracellular compartment) ^{Reference Klein15} and transient brain acidification. ^{Reference Maddock, Buonocore, Miller, Yoon, Soosman and Unruh17} Under these provocations, abnormal respiratory responses are demonstrable in people with panic disorder ^{Reference Battaglia, Ogliari, D’Amato and Kinkead18} as well as in preclinical ^{Reference D’Amato, Zanettini, Lampis, Coccurello, Pascucci and Ventura19} studies of panic pathophysiology. Specifically, it has been shown that hypercapnia induced comparable physiological changes in murine and human models of experimental panic. ^{Reference Leibold, van den Hove, Viechtbauer, Buchanan, Goossens and Lange20}

Carbon dioxide is an end-product of carbohydrate metabolism and is continuously produced within the body. The majority of bodily CO₂ is readily processed by the bicarbonate buffering system to maintain a physiological acid–base balance. CO₂ reacts with water to form carbonic acid, which in turn dissociates into hydrogen ions and bicarbonate (CO₂ + H₂O ↔ H₂CO₃ ↔ H⁺ + HCO₃ ⁻). ^{Reference Leibold, van den Hove, Esquivel, De Cort, Goossens and Strackx16} Bicarbonate buffers the hydrogen ions as otherwise these would accumulate, resulting in a decrease in pH. When CO₂ concentration increases markedly, the partial pressure of CO₂ increases rapidly, which is associated with a decrease in pH. This is followed by adaptive bodily responses such as an increase in respiratory rate to expirate excess CO₂. ^{Reference Leibold, van den Hove, Esquivel, De Cort, Goossens and Strackx16} Importantly, also in the brain, CO₂ has been shown to reduce pH. This has been demonstrated directly in rodents whereas in humans the evidence is indirect, i.e. coming from observations such as decrease in pH following intravenous bicarbonate infusion. ^{Reference Leibold, van den Hove, Esquivel, De Cort, Goossens and Strackx16}

Inhalation of air mixtures enriched with CO₂ has been developed as an experimental model to study panic attacks in the laboratory. A series of studies have demonstrated relative hyper-reactivity in panic disorder patients, good test–retest reliability, robustness against contextual manipulations and sensitivity to anti-panic medications. ^{Reference Leibold, van den Hove, Esquivel, De Cort, Goossens and Strackx16,Reference de Vos, Haj Yahya, Viechtbauer, Linden, Schruers and Leibold21} Notably, it was demonstrated that the physiological responses to excess CO₂ are similar in mice and humans, and between healthy volunteers and panic disorder patients. This allows for translation of findings in animal studies to the human situation and, ultimately, to the clinic. ^{Reference Leibold, van den Hove, Viechtbauer, Buchanan, Goossens and Lange20}

Human CO₂ hypersensitivity (i.e. acute fear and hyperventilation in response to respiratory CO₂ challenges) is substantially heritable, ^{Reference Battaglia, Ogliari, Harris, Spatola, Pesenti-Gritti and Reichborn-Kjennerud22} with 90% of genetic background shared with panic disorder and separation anxiety disorder. ^{Reference Battaglia, Pesenti-Gritti, Medland, Ogliari, Tambs and Spatola23,Reference Battaglia, Pesenti-Gritti, Spatola, Ogliari and Tambs24} Early life adversities induce gene–environment interactions that further enhance CO₂ sensitivity, ^{Reference Spatola, Scaini, Pesenti-Gritti, Medland, Moruzzi and Ogliari25} and childhood parental loss contributes to heightened human CO₂ sensitivity. ^{Reference Battaglia, Pesenti-Gritti, Medland, Ogliari, Tambs and Spatola23} Due to these features, and the reproducibility of CO₂ responses in the laboratory, the trait of CO₂ sensitivity provides a powerful preclinical context with which to study the pathophysiology underpinning panic disorder and panic attacks. ^{Reference Battaglia, Ogliari, D’Amato and Kinkead18} With many similarities to the situation of humans who experienced prolonged parental separation/loss in childhood, the repeated cross-fostering (RCF) procedure, which involves separating mice early in life from their biological mothers in favour of a rotating series of unrelated, adoptive, lactating females, induces stable, lifelong respiratory CO₂ hypersensitivity and heightened separation anxiety. ^{Reference D’Amato, Zanettini, Lampis, Coccurello, Pascucci and Ventura19,Reference Battaglia, Pesenti-Gritti, Medland, Ogliari, Tambs and Spatola23} These associations are best explained by gene–environment interaction mechanisms, ^{Reference D’Amato, Zanettini, Lampis, Coccurello, Pascucci and Ventura19} implying altered expression of certain regulatory genes. Investigation of the genetic enrichments associated with CO₂ hypersensitivity in humans and animals genome-wide yielded evidence of (a) brain-stem H3Ac/H3K4me3 histonic enrichments and upregulated messenger RNA expression of the Asic1 gene ^{Reference Cittaro, Lampis, Luchetti, Coccurello, Guffanti and Felsani26} ; (b) altered DNA methylation of the Asic2 gene concurrently in mice (brain) and humans (white blood cells) hypersensitive to CO₂ stimulation; ^{Reference Giannese, Luchetti, Barbiera, Lampis, Zanettini and Knudsen27} (c) hyperventilation in response to hypercapnia among the first generation of RCF-exposed animals, as well as among two successive generations of their normally reared offspring, through matrilineal transmission; ^{Reference Battaglia, Rossignol, Lorenzo, Deguire and Godin28} and (d) genetic variants in the ACCN2 gene, the human orthologue of the Asic1 gene, are associated with differential sensitivity to CO₂ in panic disorder patients, as well as in healthy volunteers. ^{Reference Leibold, van den Hove, Viechtbauer, Kenis, Goossens and Lange29} Parallel enhancements of ASIC1, ASIC2 and ASIC3 messenger RNA transcripts have been detected transgenerationally in the medulla oblongata and periaqueductal grey matter of RCF-lineage animals. ^{Reference Battaglia, Rossignol, Lorenzo, Deguire and Godin28} Crucially, a single, nebulised dose of the ASIC antagonist amiloride renormalised respiratory responses to CO₂ across the entire RCF lineage. ^{Reference Battaglia, Rossignol, Lorenzo, Deguire and Godin28,Reference Battaglia, Rossignol, Bachand, D’Amato and Koninck30} A single-dose pretreatment with amiloride, nebulised to reach the brain swiftly, normalised respiratory CO₂ hypersensitivity among RCF animals, thereby restoring responses to the range observed in controls, whereas RCF animals pretreated with only nebulised saline had significantly greater tidal volumes in response to CO₂. By contrast, amiloride dispensed intraperitoneally to RCF animals was not associated with efficacy in normalising the CO₂ response. A lack of efficacy similar to that seen with intraperitoneally administered amiloride was observed with saline given intraperitoneally, and with no treatment.

Because ASIC2 and ASIC1 multimerise to affect pH sensing, fear and learning, ^{Reference Wemmie, Taugher and Kreple13} these findings regarding Asic genes’ enrichment and CO₂ responses are consistent with (a) ASICs’ role in detecting brain pH changes and in evoking fear; ^{Reference Wemmie, Taugher and Kreple13} (b) functional magnetic resonance imaging evidence of brain pH changes in human panic disorder; ^{Reference Maddock, Buonocore, Miller, Yoon, Soosman and Unruh17} and (c) the associations of specific Asic1 gene single-nucleotide proteins with human panic disorder and CO₂ hypersensitivity. ^{Reference Savage, McMichael, Gorlin, Beadel, Teachman and Vladimirov31,Reference Smoller, Gallagher, Duncan, McGrath, Haddad and Holmes32} The data thus show that (a) ASICs are key pathophysiological elements of panic, as mediated by hypersensitivity to CO₂; and (b) nebulised, single-dose amiloride modulates these responses.

In summary, both genes and environment influence liability to panic attacks and panic disorder. Emotional and respiratory CO₂ hypersensitivity, a valid psychobiological marker of panic disorder/panic attacks, can be used in humans ^{Reference Battaglia, Ogliari, Harris, Spatola, Pesenti-Gritti and Reichborn-Kjennerud22} and preclinically ^{Reference Battaglia, Ogliari, D’Amato and Kinkead18,Reference D’Amato, Zanettini, Lampis, Coccurello, Pascucci and Ventura19} to translationally identify key pathophysiological mechanisms, including gene enrichment in response to environmental triggers. ASICs become overexpressed in response to environmental adversities and are integral to panic disorder/attack pathophysiology and related respiratory reactivity to hypercapnia. ^{Reference Cittaro, Lampis, Luchetti, Coccurello, Guffanti and Felsani26–Reference Battaglia, Rossignol, Lorenzo, Deguire and Godin28,Reference Battaglia, Rossignol, Bachand, D’Amato and Koninck30} The ASIC antagonist amiloride, nebulised to reach the brain, blocks hyperventilatory panic-like responses in preclinical models. ^{Reference Battaglia, Rossignol, Lorenzo, Deguire and Godin28,Reference Battaglia, Rossignol, Bachand, D’Amato and Koninck30} As research into amiloride nasal spray develops, it is anticipated that future studies will assess the impact of its administration on CO₂ challenge paradigms in humans who have experienced panic disorder or recurrent panic attacks.

Other pharmacological properties of amiloride nasal spray

Amiloride nasal spray is safe and chemically and bacteriologically stable, with models indicating rapid distribution into the human brain. ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12,Reference Azzeh, Davies, Strauss, Battaglia, Dogra and Yellepeddi33,Reference Renninger, Sayre, Battaglia, Davies, Sayre and Yellepeddi34} Extensive safety, distribution and pharmacokinetic studies of amiloride nasal spray have been completed. ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12,Reference Azzeh, Davies, Strauss, Battaglia, Dogra and Yellepeddi33} A physiologically based pharmacokinetic model ^{Reference Azzeh, Davies, Strauss, Battaglia, Dogra and Yellepeddi33} showed excellent, rapid brain concentration with maximal ASIC inhibition. Human pharmacokinetic data in 15 healthy individuals (mean age 28 years (range 21–40 years); 71% female, 65% White; 18% Hispanic/Latino, 18% Asian) ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12} demonstrated rapid absorption from the nasal mucosa, followed by a brisk plasmatic peak (ClinicalTrials.gov ID NCT04181008). Tolerability was excellent, with no grade 2+ adverse events, and the comprehensive metabolic panel (of 14 different proteins, enzymes, electrolytes, minerals and other substances), kidney profile, complete blood count with platelet count and auto-differential, estimated glomerular filtration rate, and cardiovascular parameters before and after amiloride nasal spray administration, revealed no adverse events. ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12} Amiloride nasal spray’s excellent physical, chemical and microbiological stability, extending to 90 days from preparation at room temperature and under refrigeration, have been confirmed. ^{Reference Renninger, Sayre, Battaglia, Davies, Sayre and Yellepeddi34}

Amiloride nasal spray has the potential to become a portable, rapid-acting ^{Reference Azzeh, Davies, Strauss, Battaglia, Dogra and Yellepeddi33} treatment for panic disorder, for use on an ‘as-needed’ basis only at the specific moment when a panic attack is thought to be starting. ^{Reference Yellepeddi, Battaglia, Davies, Alt, Ashby and Shipman12} Unlike some other treatments for panic attacks, ^{Reference Guaiana and Davies7} there is no evidence to suggest that amiloride has the potential to be addictive. Existing data have demonstrated that it is well tolerated and has been used safely. The body of work described herein has led to the hypothesis that, in people with panic disorder, amiloride nasal spray will block or reduce panic attack intensity, with the suggestion that it might have the capability to stop or disrupt impending panic attacks before they become full-blown attacks. The putative biological mechanisms of action described relate to channels (i.e. ASICs) that have a known physiological function (detection of extracellular pH changes) and have been characterised as existing in locations implicated in the development of panic attacks. To this extent, amiloride nasal spray can be considered a highly specific intervention moderating well-characterised targets that are directly implicated in panic attacks. This contrasts with many of the drug classes that currently have evidence of efficacy in panic attacks. In most of these, our understanding of their pharmacological action is limited to impacts on the availability of a certain neurotransmitter, without a knowledge base causally linking these changes to specific receptors or channels demonstrated to be implicated in the genesis of panic attacks. Further support for the utility of amiloride comes from animal studies illustrating its ability to normalise hypersensitive CO₂ responses in mice that experienced early life adversity.

For all of the above reasons, there is a need for clinical trials to be undertaken to test the efficacy of this preparation in individuals with panic disorder and/or recurrent panic attacks. This might include an initial comparison of the acute effects of amiloride nasal spray with a placebo, administered at the onset of panic attacks, and subsequent research into its ability to control panic attacks in comparison with existing medications and psychotherapies.