Expanding the Use of SGLT2i in Diabetes Beyond Type 2

Diabetes in totally pancreatectomized patients is notoriously difficult to manage. They not only are insulin dependent (like patients with type 1 diabetes [T1D]) but also lack pancreatic glucagon. This is compounded by severe malabsorption (partially controlled with enzyme supplementation), rapid gastric emptying (due to the asportation of the lower part of the stomach and the duodenum) and, usually, a high insulin sensitivity. All these conditions cause rapid fluctuations between hyper- and hypoglycemia (brittle diabetes), with extremely severe hypoglycemic episodes, worsened by the lack of any glucagon response.

Baekdal et al. (1), in the study published in this issue of Diabetes Care, used a standardized 3-h mixed-meal test to investigate the glucoregulatory effects of the sodium–glucose cotransporter (SGLT)2 inhibitor (SGLT2i) empagliflozin in pancreatectomized patients and matched healthy control individuals, while also evaluating the effect of SGLT2 inhibition on glucagon, other hormonal secretions, and glucose fluxes in the two groups. They found that empagliflozin, administered the day before and immediately before a standardized liquid mixed-meal test, improved postprandial glucose tolerance and, more importantly, normalized fasting glucose concentration in pancreatectomized patients—a result that is rather difficult to achieve solely with basal insulin, due to the obvious risk of hypoglycemia.

SGLT2i are a relatively new class of drugs that promote urinary glucose excretion, among other effects. The SGLT2i empagliflozin is indicated in patients with type 2 diabetes to improve blood glucose control, decrease the risk of cardiovascular adverse events, and curtail the risk of cardiovascular mortality, and, also in patients without diabetes, to reduce hospitalizations for heart failure and chronic kidney disease progression (2). Since the glucosuric effect of SGLT2i is based only on underlying blood glucose levels and kidney function (glomerular filtration rate), SGLT2i-induced glucosuria increases with hyperglycemia but decreases with approaching hypoglycemia. Further, SGLT2i (including the dual SGLT inhibitor sotagliflozin) do not inhibit renal SGLT1 (3), which becomes sufficient to reabsorb virtually all the filtered glucose when glycemia is low. That is, SGLT2i are a unique class of hypoglycemic drugs that lose all their glucose-lowering efficacy in presence of hypoglycemia. This makes them an ideal companion to insulin, as they act as a jump cushion (Fig. 1 [jump cushions are placed by firefighters to save people jumping from a burning building]), blocking the drop of glycemic levels, and favoring improved glucose control, as the study by Baekdal et al. suggests.

While this is a pivotal study in pancreatectomized patients, the jump cushion effect of SGLT2i has already been demonstrated in patients with T1D, who share the common feature of insulin deficiency. Many clinical trials using various SGLT2i in T1D have shown reduced HbA_1c and lower glycemic variability and increased time in range, as well as reductions in weight and insulin dose without a greater incidence of hypoglycemia (4,5), but at the cost of an (unacceptable?) increase of ketoacidosis episodes. Diabetic ketoacidosis (DKA) in T1D occurs when insufficient insulin is administered, e.g., when doses are skipped or substantially reduced (mostly due to fear of hypoglycemia), during illness requiring more insulin or when an insulin pump fails (6). Often, when an insulin-treated patient is started on SGLT2i, the consequent reduced glucose levels push the patient to inappropriately reduce or skip insulin doses, which becomes the main trigger leading to DKA in T1D.

However, Peters et al. (7) investigated the incidence of DKA in patients with T1D from the inTandem1 and inTandem2 trials using sotagliflozin as an adjunct to insulin. They found that the overall occurrence of DKA with sotagliflozin was <4%, (but higher than with placebo) and that implementation of a DKA mitigation plan (i.e., patient education and close monitoring) reduced DKA episodes. More importantly, the inTandem trials have shown that the use of sotagliflozin is associated with a decrease in overall hypoglycemias at all HbA_1c levels (8). In other words, the number of episodes of hypoglycemia in patients with HbA_1c of 7.5% was the same as that in patients with HbA_1c of 6.5%, thus reducing the burden of hypoglycemia in the attempt to improve glucose control. As Baekdal et al. themselves remark in their discussion, recent phase III trials have shown that the SGLT2i dapagliflozin in adjunct therapy with insulin has positive effects in patients with T1D leading to weight loss and improved glycemic levels with diminished insulin requirements. These results induced the European Medicines Agency to approve dapagliflozin in adjunct to insulin for treatment of T1D, and, although in 2021 the manufacturer withdrew the T1D indication for the drug due to fears that required changes to labeling would cause confusion among doctors prescribing it (9), it is now often used off-label for treatment of these patients. Currently, patients with T1D with heart failure or chronic kidney disease cannot be prescribed SGLT2i, even if their risk for these diseases is high.

In the abovementioned phase III trials, SGLT2i treatment was linked to a small but significant risk of DKA; however, recently in a real-life evaluation of that risk in T1D subjects over 222 patient-years of SGLT2i use investigators observed that the incidence of DKA was 0% (10). Earlier this year, in another real world study, Anson et al. (11) compared SGLT2i with glucagon-like peptide 1 receptor agonist (GLP-1 RA) therapy as adjunct to insulin in T1D and concluded that the increased risk of DKA in the SGLT2i group, and associated hospitalization, was less than the increased rate of any-cause hospitalization in the GLP-1 RA group, suggesting that SGLT2i therapy may be more beneficial as an adjunct therapy for T1D patients.

In pancreatectomized patients, the risk of DKA is even less frequent due to the absence of pancreatic glucagon secretion. This lack of pancreatic glucagon makes these patients ideal candidates for treatment with SGLT2i adjunct therapy, which stabilizes the brittle diabetes condition. In these patients, in fact, SGLT2 inhibition is useful not so much to reduce HbA_1c, already reduced by insulin, but, rather, to decrease hypoglycemia. SGLT2i normalize fasting glycemia, and their jump cushion effect makes it difficult for glycemic levels to plummet. The focus, therefore, should be not on the insulin-sparing but, rather, on the hypoglycemia-reducing effect of SGLT2 inhibitors. Failure to authorize the use of SGLT2i in such patients means not only missing out on the great advantages of glycemic stabilization and reduced risk of severe hypoglycemia but also on the significant cardio-renal benefits.