Background

While working with a celebrated pomologist in 1835, Belgian scientists Laurent-Guillaume de Koninck and Jean Servais Stas isolated phlorizin, a seemingly anodyne dihydrochalcone, from apple bark.1 About 50 years later, the German physician Josef von Mering administered phlorizin to dogs, noting its peculiar ability to stimulate polyuria and glucosuria.1 Another century would pass until the discovery of renal sodium-glucose cotransporter (SGLT) proteins, of which phlorizin is a potent and non-selective inhibitor.2 Over the past two decades, the study of SGLT type 2 (SGLT2) inhibitors has surged, illuminating multifarious benefits not only for the treatment of type 2 diabetes (T2D) but also more broadly for patients with cardiovascular-kidney-metabolic (CKM) syndrome, which includes congestive heart failure (CHF) and chronic kidney disease (CKD). Hospitalists routinely care for patients taking SGLT2 inhibitors, and the use of these agents will only increase as researchers uncover novel applications and generic formulations become widely available. This review outlines pivotal randomized controlled trials (RCTs) evaluating SGLT2 inhibitors, highlighting their clinical efficacy through pleiotropic effects, key mechanisms of action, and potential adverse effects of particular interest to hospitalists.

Methods

We performed a comprehensive PubMed search for English-language RCTs, meta-analyses, systematic and narrative reviews, observational studies, and guidelines using the search terms sodium-glucose cotransporter type 2 inhibitor, empagliflozin, dapagliflozin, canagliflozin, type 2 diabetes, congestive heart failure, and chronic kidney disease. Identified articles were evaluated manually, along with a review of their relevant references, leading to the inclusion of additional clinical trials and key publications. We prioritized high-quality original research, review articles, and guidelines, especially those with results and recommendations relevant to hospital medicine practice.

Mechanisms of Action and Cardiorenal Benefits

In hyperglycemic patients with T2D, SGLT2 inhibition (a glucose transporter expressed in the nephron’s proximal tubule responsible for ~90% of filtered glucose reabsorption) causes glucosuria, reducing serum glucose levels.3,4 The renoprotective effects of SGLT2 inhibitors, however, are more nuanced. SGLT2 inhibition prevents proximal reabsorption of glucose and sodium, increasing sodium delivery to the macula densa. This action triggers vasoconstriction of the afferent arteriole, leading to a reduction in glomerular filtration in what is termed the tubuloglomerular feedback mechanism of the juxtaglomerular apparatus.5 In the context of the glomerular hypertension and pathologic glomerular hyperfiltration that often occurs in DKD, this reduction (or normalization) in intraglomerular pressure likely accounts for the renoprotective and anti-proteinuric effects of SGLT2 inhibitors.4

SGLT2 inhibitor-induced natriuresis also benefits CHF patients by decreasing preload. Although they can lower blood pressure, SGLT2 inhibitors seem to facilitate decongestion without a high risk of hypoperfusion.4 Their cardioprotective benefits additionally stem from their proclivity to induce ketosis, which is more energetically efficient for cardiac myocytes.3 Indeed, intravenous ketone infusions have been shown to have an acutely dynamic effect on increasing cardiac output and LVEF in T2D patients with CHF.6

Other indirect mechanisms of cardiorenal protection include anti-inflammatory, antioxidant, and anti-fibrotic effects.4,7 SGLT2 inhibitors may enhance proximal tubular oxygenation by reducing energy expenditure and increasing oxygen delivery.4 The observation of increased hemoglobin and hematocrit levels (by 0.7 g/dL and 2-2.5%, respectively)8 in patients taking SGLT2 inhibitors supports the theory that these agents improve oxygen-carrying capacity, making renal and cardiac cells more resilient in the setting of acute stress.

Additional Benefits

As extensive evaluation of SGLT2 inhibitors has rapidly expanded their use in patients with CKM syndrome, researchers have recognized multiple unexpected outcomes related to other comorbidities common in this population. SGLT2 inhibitors have been shown to have a uricosuric effect as increased tubular glucose competes with uric acid for reabsorption.9 A target trial emulation study including >30,000 T2D patients found that the incidence of gout was 38% lower among patients taking SGLT2 inhibitors than among those taking sulfonylureas.10 Furthermore, there was a 66% reduction in the incidence of gout episodes requiring an emergency department visit or hospitalization. In another similarly designed study of >20,000 T2D patients with a history of nephrolithiasis, SGLT2 inhibitor initiators experienced a 33% reduction in recurrent nephrolithiasis compared with GLP-1 receptor agonist initiators (irrespective of stone composition).11 Possible explanations include increased tubular flow and favorable changes in urinary pH induced by SGLT2 inhibitors (in addition to the known potentiation of uricosuria).11

Unfortunately, hyperkalemia is a common complication for many patients with CHF or CKD receiving RAAS blockade, often leading to dose reduction or discontinuation. A retrospective cohort study of SGLT2 inhibitor initiators versus nonusers on RAAS blockade (~17% with CHF and nearly one-third with CKD stages 3-5) found that SGLT2 inhibitor initiation was associated with an 11% lower risk of hyperkalemia.12 SGLT2 inhibitor initiators were also less likely to experience RAAS blockade discontinuation (36% vs 45%). SGLT2 inhibitors increase tubular sodium delivery to the distal nephron, favoring kaliuresis via augmented epithelial sodium channel activity.

SGLT2 inhibitors have also shown promise in patients with Metabolic Dysfunction-Associated Steatohepatitis (MASH), likely through pleiotropic mechanisms involving conserved energy expenditure, anti-inflammatory and antioxidant effects, and fibrosis amelioration.13 A small placebo-controlled RCT evaluated the effect of dapagliflozin on histological hepatic outcomes in patients with biopsy-proven MASH.14 After 48 weeks, patients randomized to dapagliflozin were more likely to experience improved MASH without worsening fibrosis (53% versus 30%), resolution of MASH without worsening fibrosis (23% versus 8%), and improvement in fibrosis without worsening MASH (45% versus 20%).

Clinical Evidence

The Cardiovascular Outcome Trials (CVOTs)

Reports of adverse events related to thiazolidinediones led the U.S. Food and Drug Administration (FDA) to mandate in 2008 that all new T2D medications demonstrate cardiovascular (CV) safety.15 This necessitated the design of CV outcome trials (CVOTs) to study SGLT2 inhibitors. Although researchers sought non-inferiority, these trials yielded serendipitous results that reached beyond a narrow scope of CV risk-related parity by reporting evidence supporting the use of SGLT2 inhibitors in the setting of CKM syndrome.

CVOTs evaluating empagliflozin (EMPA-REG OUTCOME), canagliflozin (CANVAS Program), and dapagliflozin (DECLARE-TIMI 58) were published in the New England Journal of Medicine (NEJM) between 2015 and 2019.16–18 These placebo-controlled trials included nearly 35,000 patients, all with T2D and varying degrees of CV disease. Both EMPA-REG OUTCOME and the CANVAS Program demonstrated a 14% reduction in a composite CV event outcome. Additional cardiac and renal benefits also emerged, which were unexpected and set the stage for more targeted future trials. All three CVOTs reported lower rates of hospitalization for heart failure (HHF) by 27-35%, and there were reductions in a composite renal outcome for patients taking canagliflozin and dapagliflozin (by 40% and 24%, respectively). Based on these results, SGLT2 inhibitors were shown to be not only safe but efficacious. The impressive benefits occurred despite only modest improvements in hemoglobin A1C (by 0.4-0.6%), leading researchers to wonder how a weak oral hypoglycemic could so effectively impact cardiorenal outcomes (and whether such a medication could be useful in patients without T2D).

The Congestive Heart Failure Trials (CHF)

Based on the CVOTs’ persuasive evidence for cardioprotection, dapagliflozin and empagliflozin were administered to CHF patients with and without T2D across four RCTs published in NEJM between 2019 and 2022.19–22 (Table 1)

Table 1.
Cardiovascular Outcome Trials Patient Population Intervention Observation Time Primary Outcome Notable Secondary Outcomes Notable Adverse Events
EMPA-REG OUTCOME (2015)16 7020 T2D patients with established CV disease, HbA1C 7-10%, and eGFR ≥30 mL/min/1.73m2. Empagliflozin 10 mg or 25 mg versus placebo once daily. 3.1 years (median). Composite outcome (3-point MACE): death from CV causes, nonfatal MI, or nonfatal stroke.
The primary outcome occurred in 10.5% of patients in the empagliflozin group versus 12.1% in the placebo group (HR 0.86).		 Lower rate of death from CV causes in the empagliflozin group (3.7% versus 5.9%).
Lower rate of HHF in the empagliflozin group (2.7% versus 4.1%).
Lower rate of all-cause mortality in the empagliflozin group (5.7% versus 8.3%).		 Increased risk of genital mycotic infections with empagliflozin compared to placebo (6.4% versus 1.8%).
Episodes of DKA were rare (0.1% of patients or less), and there was no significant difference between groups.
CANVAS Program (2017)17 10142 T2D patients with established CV disease (65.6%) or increased risk, HbA1C 7-10.5%, and
eGFR ≥30 mL/min/1.73m2.		 Canagliflozin 100 mg or 300 mg versus placebo once daily. 3.6 years (mean). Composite outcome (3-point MACE): death from CV causes, nonfatal MI, or nonfatal stroke.
The primary outcome occurred in 26.9 (canagliflozin) versus 31.5 (placebo) patients per 1000 patient-years (HR 0.86).		 Patients taking canagliflozin experienced lower rates of HHF (HR 0.67), progression of albuminuria (HR 0.73), and a composite renal outcome including ≥40% reduction in eGFR/need for renal replacement therapy/death from renal causes (HR 0.60). Increased risk of amputation of toes, feet, or legs with canagliflozin (6.3 vs. 3.4 events per 1000 patient-years). 71% of affected patients experienced amputation no higher than the level of the metatarsal.
Increased risk of fracture with canagliflozin (15.4 versus 11.9 events per 1000 patient-years).
Increased risk of infections of male genitalia with canagliflozin (34.9 versus 10.8 events per 1000 patient-years).
Increased risk of mycotic genital infections in women with canagliflozin (68.8 versus 17.5 events per 1000 patient-years).
No significant difference in episodes of DKA (0.3-0.6 events per 1000 patient-years).
DECLARE-TIMI 58 (2018)18 17160 T2D patients with established CV disease (40.6%) or increased risk, HbA1C 6.5-12%, and CrCl ≥60 mL/min. Dapagliflozin 10 mg versus placebo once daily. 4.2 years (median). Composite outcome (3-point MACE): death from CV causes, nonfatal MI, or nonfatal stroke.
The primary outcome occurred in 8.8% of patients in the dapagliflozin group versus 9.4% in the placebo group (nonsignificant).		 Lower rate of CV death or HHF in the dapagliflozin group (4.9% versus 5.8%, primarily driven by fewer HHF).
Lower rate of a composite renal outcome (≥40% decrease in eGFR/new ESKD/death from renal or CV causes) in the dapagliflozin group (4.3% versus 5.6%).		 Increased risk of genital mycotic infections with dapagliflozin (0.9% versus 0.1%).
Increased risk of DKA with dapagliflozin, though incidence was low overall (0.3% versus 0.1%).
Decreased risk of acute kidney injury with dapagliflozin (1.5% versus 2%).
No significant differences in amputations or fractures.
Congestive Heart Failure Trials Patient Population Intervention Observation Time Primary Outcome Notable Secondary Outcomes Notable Adverse Events
DAPA-HF (2019)19 4744 HFrEF patients (average LVEF ~31%) with NYHA II-IV symptoms, elevated NTproBNP, and eGFR ≥30 mL/min/1.73m2.
41.8% of patients had T2D.		 Dapagliflozin 10 mg versus placebo once daily. 1.5 years (median). Composite outcome: worsening heart failure (hospitalization or urgent visit resulting in IV treatment) or CV death.
The primary outcome occurred in 16.3% of patients in the dapagliflozin group versus 21.2% in the placebo group (HR 0.74, consistent results regardless of T2D status).		 There was a lower rate of a first worsening heart failure event in the dapagliflozin group (10.0% versus 13.7%).
There was a lower rate of CV death in the dapagliflozin group (9.6 versus 11.5%).
There was a lower rate of all-cause mortality in the dapagliflozin group (11.6% versus 13.9%).		 No significant differences in episodes of DKA, volume depletion, hypoglycemia, amputations, or fractures.
Rates of genital mycotic infections were not reported.
EMPEROR-Reduced (2020)20 3730 HFrEF patients (average LVEF ~27%) with NYHA II-IV symptoms, elevated NTproBNP, and eGFR ≥20 mL/min/1.73m2.
49.8% of patients had T2D.		 Empagliflozin 10 mg versus placebo once daily. 1.3 years (median). Composite outcome: CV death or HHF.
The primary outcome occurred in 19.4% of patients in the empagliflozin group versus 24.7% in the placebo group (HR 0.75, consistent results regardless of T2D status).		 Patients taking empagliflozin experienced fewer HHF (388 versus 553 events).
Patients taking empagliflozin had a slower annual rate of decline in eGFR (-0.55 versus -2.28 mL/min/1.73m2 per year).		 Increased risk of genital mycotic infections with empagliflozin (1.7% versus 0.6%).
No significant differences in volume depletion, hypoglycemia, amputations, or fractures.
There were no episodes of ketoacidosis.
EMPEROR-Preserved (2021)21 5988 HFmrEF/HFpEF patients (average LVEF ~54%) with NYHA II-IV symptoms, elevated NTproBNP, and eGFR ≥20 mL/min/1.73m2.
~49% of patients had T2D.		 Empagliflozin 10 mg versus placebo once daily. 2.2 years (median). Composite outcome: CV death or HHF.
The primary outcome occurred in 13.8% of patients in the empagliflozin group versus 17.1% in the placebo group (HR 0.79, consistent results regardless of T2D status).
Results did not remain significant for patients with LVEF ≥60%.		 Patients taking empagliflozin experienced fewer HHF (407 versus 541 events).
Patients taking empagliflozin had a slower annual rate of decline in eGFR (-1.25 versus -2.62 mL/min/1.73m2 per year).		 Increased risk of genital mycotic infections with empagliflozin (2.2% versus 0.7%).
Increased risk of hypotension with empagliflozin (10.4% versus 8.6%).
No significant differences in hypoglycemia, amputations, or fractures.
There were only 9 episodes of ketoacidosis, all occurring in patients with T2D at baseline.
DELIVER (2022)22 6263 HFmrEF/HFpEF/HFimpEF patients (average LVEF ~54%) with NYHA II-IV symptoms, elevated NTproBNP, and eGFR ≥25 mL/min/1.73m2.
~45% of patients had T2D.		 Dapagliflozin 10 mg versus placebo once daily. 2.3 years (median). Composite outcome: worsening heart failure (unplanned hospitalization or urgent visit for heart failure) or CV death.
The primary outcome occurred in 16.4% of patients in the dapagliflozin group versus 19.5% in the placebo group (HR 0.82, consistent results regardless of T2D status or LVEF <60%).		 Patients taking dapagliflozin experienced fewer episodes of worsening heart failure (HR 0.79). No significant differences.
5.8% of patients in each group experienced
an adverse event leading to medication discontinuation.
Renal Trials Patient Population Intervention Observation Time Primary Outcome Notable Secondary Outcomes Notable Adverse Events
CREDENCE (2019)23 4401 albuminuric DKD patients (eGFR ≥30 mL/min/1.73m2, UACR ≥300 mg/g) with HbA1C 6.5-12% and already receiving RAASi.
Mean eGFR 56.2 mL/min/1.73m2, median UACR 927 mg/g.		 Canagliflozin 100 mg versus placebo once daily. 2.6 years (median).
Trial was stopped early given clear benefits.		 Composite outcome: ESKD (dialysis, transplantation, or a sustained estimated GFR of <15 ml per minute per 1.73 m2), a doubling of the serum creatinine level, or death from renal or CV causes.
The primary outcome occurred in 43.2 (canagliflozin) versus 31.1 (placebo) patients per 1000 patient-years (HR 0.70).		 Patients taking canagliflozin experienced lower risks of ESKD (HR 0.68), MACE (HR 0.80), and HHF (HR 0.61). There was a 10-fold increased risk of DKA for patients taking canagliflozin compared to placebo, though the absolute risk was low (2.2 vs. 0.2 per 1000 patient-years).
No significant differences in the risk of lower-limb amputation or fractures.
DAPA-CKD (2020)24 4304 albuminuric CKD patients (eGFR 25-75 mL/min/1.73m2, UACR 200-5000 mg/g) already receiving RAASi (if tolerated, >97% of patients).
Mean eGFR ~43 mL/min/1.73m2, median UACR ~950 mg/g.
~67% of patients had T2D.		 Dapagliflozin 10 mg versus placebo once daily. 2.4 years (median).
Trial was stopped early given clear benefits.		 Composite outcome: sustained decline in the eGFR of at least 50%, ESKD, or death from renal or CV causes.
The primary outcome occurred in 9.2% of patients taking dapagliflozin versus 14.5% of patients taking placebo (HR 0.61, consistent results regardless of T2D status).		 Patients taking dapagliflozin experienced lower risks of a composite of death from CV causes or HHF (HR 0.71) and all-cause mortality (HR 0.69). Increased risk of hypovolemia with dapagliflozin (5.9% versus 4.2%).
No significant differences in hypoglycemia, episodes of DKA, amputations, or fractures.
EMPA-KIDNEY (2022)25 6609 CKD patients (eGFR ≥20-<45 mL/min/1.73m2 or eGFR ≥45-<90 with UACR ≥200 mg/g) already receiving RAASi (if tolerated, ~85% of patients).
Mean eGFR ~37 mL/min/1.73m2, median UACR ~330 mg/g.
~46% of patients had T2D.		 Empagliflozin 10 mg versus placebo once daily. 2.0 years (median).
Trial was stopped early given clear benefits.		 Composite outcome: progression of kidney disease (defined as ESKD, a sustained decrease in eGFR to <10 ml per minute per 1.73 m2, a sustained decrease in eGFR of ≥40% from baseline, or death from renal causes) or death from CV causes.
The primary outcome occurred in 13.1% of patients taking empagliflozin versus 16.9% of patients taking placebo (HR 0.72, consistent results regardless of T2D status).
Patients with UACR >300 mg/g experienced the most benefit.		 The rate of hospitalization for any cause was lower for patients taking empagliflozin (HR 0.86). There were numerically more episodes of DKA (6 versus 1) and lower limb amputations with empagliflozin (28 versus 19), though these results were nonsignificant.
DKA occurred in one patient taking empagliflozin who did not have T2D at baseline.

Note. CKD = chronic kidney disease; CrCl = creatinine clearance; CV = cardiovascular; DKA = diabetic ketoacidosis; DKD = diabetic kidney disease; eGFR = estimated glomerular filtration rate; ESKD = end stage kidney disease; HbA1C = hemoglobin A1C; HFimpEF = heart failure with improved ejection fraction; HFmrEF = heart failure with mildly reduced ejection fraction; HFpEF = heart failure with preserved ejection fraction; HFrEF = heart failure with reduced ejection fraction; HHF = hospitalization for heart failure; HR = hazard ratio; IV = intravenous; LVEF = left ventricular ejection fraction; MACE = major adverse cardiovascular events; NTproBNP = N-terminal pro-B-type natriuretic peptide; NYHA = New York Heart Association; RAASi = renin-angiotensin-aldosterone system inhibitors; T2D = Type 2 Diabetes; UACR = urine albumin-creatinine ratio.

DAPA-HF (dapagliflozin) and EMPEROR-Reduced (empagliflozin) included patients with heart failure with reduced ejection fraction (HFrEF, left ventricular ejection fraction (LVEF) <40%, average LVEF ~30%). Composite outcomes, including CV death and HHF, occurred about 25% less frequently in the SGLT2 inhibitor groups. These results remained significant regardless of T2D status, laying the groundwork for a new pillar of guideline-directed medical therapy for HFrEF. EMPEROR-Preserved (empagliflozin) and DELIVER (dapagliflozin), which evaluated CHF patients with LVEF >40% (average LVEF ~54%), similarly demonstrated decreases in CV death and HHF by about 20% – again, irrespective of a comorbid T2D diagnosis. Although effective treatments for heart failure with preserved ejection fraction (HFpEF) had largely remained elusive throughout the first two decades of the 21st century, these trials have dramatically changed the landscape of HFpEF management.

Initiating SGLT2 inhibitors as quickly as possible for appropriate CHF patients is an important consideration for hospitalists, as cardioprotective benefits start to appear after as little as two weeks of treatment. In EMPEROR-Reduced and DELIVER, for example, statistically significant differences in composite outcomes occurred at 12 days and 13 days after randomization, respectively.26,27 On the other hand, these benefits are likely attenuated within 30 days of SGLT2 inhibitor discontinuation, as illustrated by a secondary analysis of EMPEROR-Reduced and EMPEROR-Preserved.28

Several studies have evaluated the efficacy and safety of initiating SGLT2 inhibitors in the inpatient setting for CHF management, highlighting the pivotal role hospitalists can play in ensuring the addition of this essential medication class. The EMPULSE Trial randomized >500 inpatients with newly diagnosed or acute on chronic decompensated CHF to empagliflozin versus placebo (started when participants were clinically stable, median time three days after hospital admission).29 After treatment for up to 90 days, patients receiving empagliflozin were more likely to experience clinical benefit (a composite of mortality, heart failure events, time to first heart failure event, and symptom burden). The significant clinical benefit (stratified win ratio 1.36) was consistent regardless of baseline T2D, LVEF, or de novo versus chronic CHF. No episodes of DKA were reported, and patients taking empagliflozin had numerically fewer episodes of acute renal failure (7.7% versus 12.1%). Another study, DAPA ACT HF-TIMI 68, reported results from a meta-analysis of three RCTs (including the EMPULSE Trial) evaluating in-hospital SGLT2 inhibitor initiation for CHF patients.30 Collecting data from >3,500 patients, the authors observed that inpatient introduction of SGLT2 inhibitors was associated with a reduction in early risk of CV death or worsening CHF (hazard ratio (HR) 0.71) and in all-cause mortality (HR 0.57) during the post-discharge period.

The Renal Trials

Following in the footsteps of the CVOTs, three placebo-controlled RCTs published in NEJM between 2019 and 2022 provided strong evidence for even more expanded use of SGLT2 inhibitors by demonstrating benefits in patients with CKD.23–25 (Table 1)

CREDENCE (canagliflozin) evaluated patients with T2D and CKD with albuminuria, defined by an estimated glomerular filtration rate (eGFR) of 30-<90 mL/min/1.73m2 and by a urine albumin-creatinine ratio (UACR) of 300-5,000 mg/g. All patients were receiving treatment with maximally tolerated renin-angiotensin-aldosterone system (RAAS) blockade, the standard of care at the time for diabetic kidney disease (DKD) with albuminuria. The risk of a composite outcome (end-stage kidney disease, doubling of creatinine, or death from renal or CV causes) was found to be 30% lower in the canagliflozin group. The DAPA-CKD (dapagliflozin) results extended the renoprotective benefits of SGLT2 inhibitors beyond T2D. This trial included patients with CKD (eGFR 25-75 mL/min/1.73m2) with albuminuria (UACR 200-5,000 mg/g) who were already on maximally tolerated RAAS blockade (if tolerated). Approximately one-third of enrolled patients did not have T2D. The dapagliflozin group ultimately experienced a 44% lower risk of a composite renal outcome. EMPA-KIDNEY (empagliflozin) reported a 28% lower risk of CKD progression or death from CV events in patients with not only a wider range of CKD etiologies (only about one-third attributed to DKD), but also with greater breadth in terms of proteinuric burden: patients with an eGFR of 20-44 mL/min/1.73m2 were eligible for enrollment regardless of the presence of albuminuria, and patients with an eGFR of 45-89 mL/min/1.73m2 were included if their UACR was at least 200 mg/g.

Notably, all three renal trials were terminated early after interim analyses demonstrated convincing clinical efficacy, significantly increasing the population of patients that could benefit from SGLT2 inhibitors. More recently, there has been increasing interest in the potential for SGLT2 inhibitor-related cardiorenal benefits in patients with more severe CKD, including those on hemodialysis and status post renal transplant (one trial evaluating this population, the RENAL LIFECYCLE Trial, is ongoing).31

Adverse Effects

Although the risk of serious adverse events reported in the landmark RCTs was repeatedly lower for SGLT2 inhibitors than for placebo,7 this speaks more to their clinical efficacy than overall safety. Hospitalists must recognize several important complications related to this medication class.

Volume depletion and hypotension are understandable worries before starting SGLT2 inhibitors given their diuretic effect. Any clinically significant excess risk is likely minimal without tenuous volume status or unstable hemodynamics.7 Although it is common to observe an acute decline in eGFR following the initiation of an SGLT2 inhibitor, this anticipated hemodynamic effect is transient and does not typically reflect an acute kidney injury (AKI). Observational and meta-analysis data have not revealed an increased risk of AKI associated with SGLT2 inhibitor use.32,33 Furthermore, there is emerging evidence that SGLT2 inhibitors are in fact associated with a reduced risk of acute renal complications: one meta-analysis reported a lower risk of AKI with SGLT2 inhibitors,34 and a recent RCT in patients undergoing elective cardiac surgery observed a lower incidence of AKI in the seven days following cardiac surgery for patients who started dapagliflozin one day prior to surgery compared with placebo.35

While the energetic shift towards ketosis contributes to SGLT2 inhibitors’ efficacy, there is an increased risk of ketoacidosis, including euglycemic ketoacidosis (which requires a high index of clinical suspicion given the possibility of a normal serum glucose level).4 Although the absolute risk of SGLT2 inhibitor-induced ketoacidosis is low (<.5%),36 experts recommend avoiding their use in patients with type 1 diabetes, ketosis-prone T2D, or in the presence of other risk factors for ketoacidosis (e.g., BMI ≤ 25 kg/m2, HbA1C >10%, fasting state, acute infection).4,36–38 Importantly, SGLT2 inhibitor-induced ketoacidosis theoretically should not occur in patients without T2D,7 and thus there are many scenarios in which inpatient continuation or initiation of an SGLT2 inhibitor is not unreasonable. Uninterrupted SGLT2 inhibitor administration to appropriately selected inpatients with T2D can also be considered, as a 2024 meta-analysis including >35,000 T2D patients reported that continuing baseline SGLT2 inhibitors was associated with a lower rate of in-hospital mortality and mildly decreased length of stay without increasing the risk of AKI.39

Hospitalists must also recognize that as a consequence of glucosuria, SGLT2 inhibitors are associated with a substantially increased risk of genital mycotic infections compared to placebo in patients with T2D (risk ratio 3.65).40 Educating patients about perineal hygiene has been shown to decrease the risk of such infections in this population.41 Although exceedingly rare, necrotizing fasciitis of the perineum, also known as Fournier Gangrene, is a potentially catastrophic complication of SGLT2 inhibitors, leading the U.S. FDA to issue a safety warning in 2018.42 Although RCTs comparing SGLT2 inhibitors with placebo have not shown a consistent association with bacterial urinary tract infections (UTIs), the U.S. FDA issued a safety warning on the risk of serious UTIs.40 In a retrospective study, however, SGLT2 inhibitor discontinuation after a bacterial UTI was associated with worse CV/renal outcomes and increased CV mortality without decreasing the risk of recurrent infections.43

Uncertainty also remains about other adverse effects related to SGLT2 inhibitors. Although the CANVAS Program (canagliflozin) reported an increased risk of predominantly toe/metatarsal amputations, such an association was not clearly shown in RCTs involving other SGLT2 inhibitors.7,17 The CANVAS Program also reported an increased risk of bone fractures, hypothesized to have been a consequence of volume depletion-related falls; this finding, however, was not confirmed in any of the other RCTs involving SGLT2 inhibitors.17,44,45

Limitations

Although the aforementioned RCTs were undoubtedly rigorous, including more than 70,000 patients and contributing to a modern revolution in the treatment of CKM syndrome, several important limitations remain. It should be recognized that major pharmaceutical companies sponsored all the RCTs. The trial populations were also highly selected, excluding many patients often treated by hospitalists (e.g., those with severe renal impairment or recent cardiovascular events, and those unable to attend frequent outpatient appointments reliably). This may limit the generalizability of the RCTs’ findings to the complex, heterogeneous population of patients routinely encountered on any hospital medicine service.

Conclusion

From the serendipitous discovery of phlorizin nearly two centuries ago to the surge in RCTs over the past two decades, SGLT2 inhibition has yielded a wide range of benefits for human health. SGLT2 inhibitors have evolved from oral hypoglycemics for T2D into practice-changing therapies for CHF and CKD. Hospitalists will undoubtedly care for increasing numbers of patients who may benefit from SGLT2 inhibitors as well as those already receiving therapy at baseline. Careful patient selection, attention to acute clinical status, and consideration of the most common adverse effects are essential to their safe administration in the inpatient setting. Hospitalists can play an important role not only in ensuring appropriate initiation and continuation of SGLT2 inhibitors, but also in preventing avoidable complications resulting from indiscriminate use in hospitalized patients (Table 2). The convincing clinical benefits of SGLT2 inhibitors for patients with CKM syndrome make familiarity with their mechanisms of action and potential risks increasingly important for hospitalists.

Table 2.4 Practical SGLT2 Inhibitor Questions for Hospitalists
4 Practical SGLT2 Inhibitor Questions for Hospitalists
  1. Which patients may benefit from an SGLT2 inhibitor?
  • Patients with T2D (either as initial or add-on therapy), especially in the setting of comorbid CHF and/or DKD.
  • Patients with HFrEF or HFpEF (regardless of the presence of T2D).
  • Patients with CKD and an eGFR ≥20 mL/min/1.73m2 (regardless of the presence of T2D), especially those with albuminuria (UACR ≥200 mg/g).
  • Consider additional benefits of SGLT2 inhibitors for patients with CKM syndrome and other comorbidities including gout, nephrolithiasis, hyperkalemia related to RAAS blockade, and MASH.
  1. What caveats should be kept in mind before starting an SGLT2 inhibitor?
  • Patients with tenuous volume status or unstable hemodynamics are at increased risk of hypotension and renal hypoperfusion injury related to SGLT2 inhibitors.
  • Patients with type 1 diabetes, ketosis-prone T2D, or other risk factors for ketoacidosis (e.g., BMI ≤ 25 kg/m^2^, HbA1C >10%, fasting state, acute infection) are at increased risk of SGLT2 inhibitor induced ketoacidosis, including euglycemic ketoacidosis.
  • A history of severe or frequent genital mycotic infections should prompt a risks/benefits discussion before SGLT2 inhibitor initiation.
  1. Should hospitalists always hold a patient’s SGLT2 inhibitor on admission?
  • Any decision around holding or continuing a hospitalized patient’s SGLT2 inhibitor should be individualized with special attention paid to the indication, comorbidities, and acute clinical picture (e.g., risk of ketoacidosis, NPO status, tenuous volume status).
  • Given the risk of ketoacidosis, the American Diabetes Association recommends holding SGLT2 inhibitors in T2D patients for 3-4 days before elective surgeries.46
  • Remember that SGLT2 inhibitor induced ketoacidosis is not expected to occur in patients without T2D. Thus, it is reasonable to continue SGLT2 inhibitors in the inpatient setting for non-diabetic patients with CHF/CKD and stable volume status.
  • Keeping in mind comorbid CHF or CKD in addition to the risks of ketoacidosis, hypovolemia, and planned surgeries, SGLT2 inhibitors can also be continued for many inpatients with T2D.
  1. How quickly do the beneficial effects from an SGLT2 inhibitor take effect, and how quickly do they dissipate?
  • CV/CHF benefits start to appear after as little as two weeks of treatment with SGLT2 inhibitors. Hospitalists should thus recommend initiating SGLT2 inhibitors in appropriate patients before discharge rather than deferring to the outpatient setting.
  • CV/CHF benefits of SGLT2 inhibitors are likely attenuated within 30 days of discontinuation. Hospitalists must communicate clearly with patients and outpatient providers if an SGLT2 inhibitor is held at time of discharge to prevent inadvertent permanent discontinuation.

Disclosures/Conflicts of Interest

None

Corresponding author

Sam Woodworth, MD
Clinical Assistant Professor in the Department of Medicine
Tufts University School of Medicine
MaineHealth Hospital Medicine
22 Bramhall Street
Portland, ME 04102
samuel.woodworth@mainehealth.org
Phone: 207-662-4618 | fax: 207-662-6254