Sweet success? SGLT2 inhibitors and diabetes

What an exciting time this is for diabetic therapeutics! In addition to insulin, sodium-glucose cotransporter 2 (SGLT2) inhibitors have emerged as a new and promising therapeutic approach for lowering blood glucose level. SGLT2 inhibitors almost seem too good to be true: not only do they lower blood glucose levels in Type 2 diabetes patients, they also lead to weight loss, lower blood pressure, and reduce cardiovascular mortality (9, 16, 17, 19). Currently, three SGLT2 inhibitors have been approved in the United States (dapagliozin, canagliozin, and empagliozin). Are these SGLT2 inhibitors really “miracle drugs” that have brought us sweet success? What might the long-term side effects be? We highlight a few recent advances in our understanding of the effect of SGLT2 inhibition on the diabetic kidney. A comprehensive review of renal glucose handling can be found in a recent review (14). The mechanism of action and efficacy of SGLT2 and their inhibitors have been examined in other reviews (1, 18).

Inhibiting SGLT2 lowers proximal tubule uptake of not only glucose but Na⁺ as well, thereby shifting Na⁺ transport to downstream nephron segments. Given that the transport efficiency of those downstream segments is lower than the proximal tubule (3), a critical question is whether SGLT2 inhibition has the detrimental effect of lowering medullary O₂ tension. OʼNeill et al. (10) sought insight into this question by measuring renal hemodynamics and renal O₂ homeostasis in diabetic rats before and after acute SGLT inhibition using phlorizin. They found that SGLT inhibition improved cortical O₂ tension but reduced medullary O₂ tension.

OʼNeill et al.ʼs experimental observations were confirmed by a theoretical study (6) by my group, in which we used a computational model of the rat nephron (5, 7, 8) to simulate solute transport and tubular metabolism in untreated diabetes and under SGLT2 inhibition. Why reproduce experimental findings using computer simulations?” you might ask. Computer simulations can reveal mechanisms that underlie experimental observations (4, 13). In this case, modeling results point to the redistribution by acute SGLT2 inhibition of active Na⁺ transport to less efficient nephron segments, such as the medullary thick ascending limb. The result is a reduction in medullary O₂ tension, as seen in Ref. 6.

Chronic SGLT2 inhibition experiments more closely mimic diabetic therapeutics using a SGLT2 inhibitor. How is medullary oxygenation affected by chronic administration of SGLT2 inhibitors? Model simulations predicted that chronic SGLT2 blockade in diabetes lowers cortical O₂ consumption and raises medullary O₂ consumption, particularly in S3 segments (6). Given that the S3 segment is known to be particularly vulnerable to hypoxic injury, the predicted increase in its O₂ consumption suggests that SGLT2 inhibitors may increase hypoxia in that segment. This is a particular concern in a diabetic kidney, in light of the observation that peritubular capillary flow decreases in diabetes (2). On the other hand, the reduction in O₂ tension may stimulate hypoxia-inducible factor-2 and enhance erythropoietin release from interstitial cells. The resulting increase in hematocrit (12) may improve not only the oxygenation of the outer medulla, but O₂ delivery to other organs as well. Indeed, some studies have indicated renoprotective effects of SGLT2 inhibitors (15).

Thus, it is somewhat surprising and alarming that a number of cases of acute kidney injury (AKI) have been reported for canagliozin and dapagliozin, with some patients requiring dialysis. As a result, AKI is now listed as a potential side effect of SGLT2 inhibitors, likely resulting in some diabetic patients with early-stage chronic kidney disease not being prescribed SGLT2 for fear of AKI. Saly and Perazella (11) discussed this issue and recognized that the potential mechanisms for AKI include uric acid toxicity and increased downstream glucose reabsorption by the SGLT1 transporter in the S3 segment. They argued the importance of ascertaining whether the reported AKI represents true structural kidney injury or a functional decline in glomerular filtration rate. The authors proposed combining readily available clinical tools (urine microscopy), experimental biomarkers of kidney injury, and kidney-on-a-chip technology to assess the nephrotoxicity of SGLT2 inhibitors (11).

It is clear that more studies are needed to better understand the long-term effects of SGLT2 inhibitors, the mechanisms underlying their cardiovascular protective effect, and the extent to which that protective effect and the underlying mechanisms may change in patients with impaired kidney function and reduced glomerular filtration rate.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.T.L. conceived and designed research; A.T.L. analyzed data; A.T.L. interpreted results of experiments; A.T.L. drafted manuscript; A.T.L. edited and revised manuscript; A.T.L. approved final version of manuscript.