SGLT2 Inhibitors in HFpEF with CKD: What the Evidence Shows

The Clinical Question

A 72-year-old woman with heart failure with preserved ejection fraction (LVEF 58%), stage 3b CKD (eGFR 34 mL/min), type 2 diabetes, and a BMI of 31 presents for follow-up. She takes lisinopril 20 mg, amlodipine 5 mg, furosemide 40 mg daily, and metformin 500 mg twice daily (reduced for renal function). Her BNP is 280 pg/mL. She had one heart failure hospitalization in the past 12 months.

Should she start an SGLT2 inhibitor? Which one, at what dose? And what does the evidence actually show for patients with her specific combination of HFpEF and moderate-to-severe CKD — not the average trial participant, but someone with these numbers?

This question sits at the intersection of three major trial programs, multiple guideline updates, and a rapidly evolving evidence base. Here is what the data show when you look at it through the lens of this patient.

The Landmark Trials

EMPEROR-Preserved (Empagliflozin)

Published in The New England Journal of Medicine in August 2021, EMPEROR-Preserved was the trial that cracked HFpEF open for SGLT2 inhibitors. It enrolled 5,988 patients with LVEF >40% and NYHA Class II-IV symptoms. Empagliflozin reduced the primary endpoint — cardiovascular death or heart failure hospitalization — by 21% (HR 0.79, 95% CI 0.69-0.90, p < 0.001).

For our patient, three features of this trial matter most:

• Renal subgroup data. Among patients with eGFR below 60 mL/min (n=2,112), the treatment effect held (HR 0.78, 95% CI 0.64-0.95). For eGFR below 45 (n=731), the point estimate remained favorable (HR 0.81) though the confidence interval widened (0.60-1.09) with the smaller sample. The signal is consistent even as you move down the eGFR curve.
• Diabetes subgroup. Approximately 49% of enrolled patients had type 2 diabetes. The treatment effect held regardless of diabetes status (interaction p = 0.41). In the diabetic subgroup, empagliflozin reduced the primary endpoint by 22% (HR 0.78, 95% CI 0.65-0.93).
• eGFR decline. Empagliflozin slowed the annual rate of eGFR decline by 1.36 mL/min/1.73m2/year compared to placebo (p < 0.001). For a patient with eGFR 34 who is approaching the threshold where treatment options narrow further, this renal protection matters as much as the heart failure benefit.

DELIVER (Dapagliflozin)

Published in The New England Journal of Medicine in September 2022, DELIVER enrolled 6,263 patients with LVEF >40%. Dapagliflozin reduced the composite of cardiovascular death or worsening heart failure by 18% (HR 0.82, 95% CI 0.73-0.92, p < 0.001).

Key data for patients with comorbid CKD:

• Renal inclusion criteria. DELIVER included patients with eGFR as low as 25 mL/min — a lower threshold than EMPEROR-Preserved (which required eGFR 20 or above). More data for patients with advanced CKD.
• eGFR subgroups. In patients with eGFR below 60 (n=2,830), HR 0.78 (95% CI 0.66-0.92). For eGFR below 45 (n=1,083), HR 0.82 (95% CI 0.63-1.07). Point estimates consistently favorable across renal strata, with widening confidence intervals in the lowest eGFR groups — reflecting smaller sample sizes, not a loss of treatment effect.
• Combined endpoint with renal outcomes. A prespecified analysis including a composite renal endpoint (sustained 50% or greater eGFR decline, end-stage kidney disease, or renal death) showed HR 0.71 (95% CI 0.49-1.02) — not statistically significant in isolation, but directionally consistent with renal protection.

DAPA-CKD (Dapagliflozin — Renal Focus)

DAPA-CKD (published in The New England Journal of Medicine, October 2020) was not a heart failure trial, but it enrolled 4,304 patients with CKD (eGFR 25-75, UACR 200-5000 mg/g), including patients with and without diabetes and a significant proportion with concurrent heart failure. For the HFpEF-CKD overlap population, this trial provides the strongest renal data.

• Primary renal endpoint. Dapagliflozin reduced the composite of sustained 50% or greater eGFR decline, ESKD, or renal/CV death by 39% (HR 0.61, 95% CI 0.51-0.72, p < 0.001). This is the most robust renal protection signal of any SGLT2 inhibitor trial.
• Heart failure hospitalization (secondary endpoint). Among all DAPA-CKD patients, HF hospitalization dropped 29% (HR 0.71, 95% CI 0.55-0.92). In the subgroup with baseline heart failure (n=468), the effect was even more pronounced (HR 0.64, 95% CI 0.43-0.95). Dual benefit in one drug.
• Low eGFR subgroup. In patients with baseline eGFR 25-45 (n=1,398), dapagliflozin reduced the primary endpoint by 36% (HR 0.64, 95% CI 0.49-0.83). This is directly relevant to our patient with eGFR 34.

Meta-Analytic Data

A 2023 meta-analysis by Vaduganathan et al. in The Lancet pooled data from EMPEROR-Preserved, DELIVER, EMPEROR-Reduced, and DAPA-HF (n=21,947). Among HFpEF patients specifically (LVEF >40%), SGLT2 inhibitors reduced the composite of CV death or HF hospitalization by 20% (HR 0.80, 95% CI 0.73-0.87). The benefit was consistent across eGFR subgroups, diabetes status, age, sex, and NYHA class. No subgroup was left behind.

A separate pooled analysis by Neuen et al. in the Journal of the American Society of Nephrology (2024) focused on renal outcomes across all major SGLT2 inhibitor trials. In patients with baseline eGFR 25-45, SGLT2 inhibitors reduced kidney disease progression by 37% (HR 0.63, 95% CI 0.53-0.75) — a larger relative risk reduction than in patients with higher baseline eGFR. The renal benefit may be greatest in the patients who need it most.

Safety Considerations in CKD

The safety profile of SGLT2 inhibitors in moderate CKD warrants specific attention for our patient with eGFR 34.

The Initial eGFR Dip

SGLT2 inhibitors cause a hemodynamic reduction in eGFR of approximately 3-5 mL/min in the first 2-4 weeks. This is not nephrotoxicity — it reflects reduced intraglomerular pressure via tubuloglomerular feedback, the same mechanism that mediates the long-term renal protection. In DAPA-CKD, the initial dip was fully reversible upon discontinuation and was followed by a significantly slower rate of eGFR decline compared to placebo. The short-term dip protects the long-term trajectory.

But here is the practical wrinkle: for a patient starting at eGFR 34, a 5 mL/min dip brings her to 29 — crossing the eGFR 30 threshold that triggers prescribing restrictions for other medications (notably metformin, which many guidelines recommend discontinuing below eGFR 30). You need to anticipate this dip, warn the patient, and plan for temporary metformin adjustment.

Glycemic Efficacy and DKA Risk

At eGFR 34, the glucose-lowering efficacy of SGLT2 inhibitors is substantially reduced — less glucose is filtered and available for inhibition. The HbA1c reduction at eGFR 30-45 is approximately 0.2-0.3%, compared to 0.7-0.9% at normal GFR. Do not start an SGLT2 inhibitor at this eGFR for glycemic control — the indication here is cardiorenal protection.

Euglycemic DKA risk is low but not zero, particularly in patients with type 2 diabetes on concurrent insulin or sulfonylureas. For our patient on metformin only, this risk is minimal.

Volume Status and Diuretic Interactions

Our patient takes furosemide 40 mg daily. SGLT2 inhibitors have a mild osmotic diuretic effect. The combination can lead to volume depletion, particularly in elderly patients. Practical management: consider reducing furosemide to 20 mg when initiating the SGLT2 inhibitor, then reassess volume status at 1-2 weeks. Monitor weight, blood pressure, and symptoms of orthostatic hypotension.

Current Guideline Positions

The 2023 ACC/AHA/HFSA Focused Update gives SGLT2 inhibitors a Class I recommendation for patients with HFpEF (LVEF >40%) to reduce HF hospitalization and CV death, regardless of diabetes status. The 2024 KDIGO guidelines recommend SGLT2 inhibitors for patients with CKD and eGFR 20 or above with either albuminuria or heart failure. Both guideline sets support initiating therapy in our patient.

Empagliflozin versus dapagliflozin? Guidelines do not clearly differentiate. Both have Class I evidence for HFpEF. Dapagliflozin has a lower eGFR enrollment threshold in its CKD trial (DAPA-CKD enrolled down to eGFR 25; EMPA-KIDNEY enrolled down to eGFR 20). In practice, either agent is reasonable, with the choice often determined by formulary availability and insurance coverage.

Synthesizing the Evidence for This Patient

For our 72-year-old with HFpEF, eGFR 34, type 2 diabetes, and a recent HF hospitalization:

• The evidence strongly supports initiation. Multiple large RCTs and meta-analyses show consistent benefit for both heart failure and renal outcomes in patients with this profile. You would need to treat 21 patients for two years to prevent one HF hospitalization based on pooled HFpEF data — a number most physicians would find compelling for a well-tolerated daily pill.
• No dose adjustment needed. Both empagliflozin 10 mg and dapagliflozin 10 mg are used at full dose regardless of eGFR (down to eGFR 20 per current labeling). Unlike many other medications in CKD, SGLT2 inhibitors do not require renal dose adjustment.
• Expect and plan for the eGFR dip. Recheck creatinine and eGFR at 2-4 weeks. An eGFR decline to approximately 29-31 is expected and should not trigger discontinuation. Metformin may need to be held temporarily if eGFR falls below 30, then reassessed once eGFR stabilizes.
• Adjust the diuretic. Reduce furosemide to 20 mg when starting the SGLT2 inhibitor. Reassess volume status at 1-2 weeks.
• Dual benefit trajectory. Over 12-24 months, expect stabilization or slowing of eGFR decline (approximately 1-2 mL/min/year slower than without the drug) and reduced heart failure hospitalization risk. Both outcomes are supported by RCT data in her eGFR range.

The kind of question this patient raises — spanning cardiology, nephrology, and endocrinology, requiring subgroup-level data from multiple trials — is exactly where having access to a tool that synthesizes evidence across specialties and surfaces the specific data points for a specific patient profile saves time and improves decisions. Ailva handles these complex, cross-system clinical questions. See how Ailva handles multi-system cases.

For more on why this cross-specialty approach matters, read beyond single-specialty answers: why cross-system reasoning matters. And for the foundational challenge that tools like Ailva address, see the bench-to-bedside gap: why evidence takes 17 years to reach patients.