The mechanism of action of sodium–glucose co‐transporter 2 inhibitors is similar to carbonic anhydrase inhibitors

Abstract

Butler et al.,1 in their review about the potential role of sodium–glucose co-transporter-2 inhibitors (SGLT2i) in the treatment of heart failure (HF), do not mention the carbonic anhydrase (CA) inhibitors, whose mechanism of action has remarkable similarities with SGLT2i. Although there are no studies in humans,2 it is established that around 30% of the filtered Na+ is reabsorbed through the Na+/H+ isoform 3 (NHe3) exchanger, which is located in the proximal convoluted tubule (PCT), whereas up to 15% of tubular Na+ is reabsorbed through SGLT2 in situations of hyperglycaemia. The SGLT2i inhibit both the SGLT2 and the NHe3 exchangers that share the microdomain with SGLT2,2 and they have a natriuretic effect independent of the glycaemic level.3 On the other hand, NHe3 is activated in HF and is involved in the states of resistance to diuretics.4 Carbonic anhydrase is an enzyme located on the luminal membrane of PCT cells that accelerates the formation of H2CO3 and its subsequent conversion into H2O and CO2 using H+ and HCO3−, which favours the mobilization of NHe3 substrates and increases the tubular reabsorption of Na+. CA inhibitors, such as acetazolamide, act to inhibit the CA, slowing down the H2CO3 circuit and decreasing NHe3 activity, leading to natriuresis.5 The SGLT2i and CA inhibitors indirectly inhibit NHe3, cause mild to moderate natriuresis and bicarbonaturia, and activate the juxtaglomerular system by decreasing hyperfiltration and proteinuria, secondary to increased Na+ and Cl− reabsorption by macula densa (MD). This reabsorption would activate the juxtaglomerular system allowing the local secretion of adenosine that in turn stimulates the afferent arteriole vasoconstriction, which leads to a decrease in hyperfiltration. This haemodynamic fact does not occur with loop diuretics, thiazides, or K+-sparing diuretics. The loop diuretics block the carrier Na+/K+/2Cl− in the MD, and the others act beyond the ascending loop of Henle and MD, so these drugs would have no proteinuric or nephroprotective effect.5 There are differences in the natriuretic effect between SGLT2i and CA inhibitors. CA inhibitors lose their diuretic effect in the long term due to compensatory mechanisms of distal reabsorption.5 However, the SGLT2i maintain their long-term diuretic effect. Thus, currently, the clinical use of acetazolamide in HF is restricted to patients with diuretic resistance. In contrast, the diuretic effect of SGLT2i is maintained over time, probably due to the glucosuric effect6 that modifies renin levels.1 Despite these differences in their long-term temporal effect considering the similarities in the mechanism of action of SGLT2i and CA inhibitors, and considering the benefit demonstrated by SGLT2i, it would be interesting to develop specific clinical trials with CA inhibitors in patients with diabetes and HF (https://clinicaltrials.gov/ct2/show/NCT01973335).