Sodium-Glucose Cotransporter 2 Inhibitors in Heart Failure with Preserved Ejection Fraction: A Systematic Review and Meta-Analysis with Trial Sequential Analysis

This systematic review and meta-analysis of 59 randomized controlled trials demonstrates that SGLT2 inhibitors significantly reduce all-cause mortality and heart failure hospitalizations across the full spectrum of heart failure, supporting their role as a foundational therapy despite an increased risk of genital infections.

The Gist

Imagine your heart is like the engine of a car. Sometimes, that engine gets worn out and struggles to pump enough fuel (blood) to the rest of the car. This condition is called Heart Failure. For a long time, mechanics (doctors) had a few tools to fix the engine, but they weren't perfect, especially for certain types of wear and tear.

Recently, a new type of tool appeared: SGLT2 inhibitors. Originally, these were designed to help people with diabetes lower their blood sugar. But doctors noticed something surprising: they also seemed to help the heart engine run much better.

This paper is a massive "report card" that gathered data from 59 different scientific experiments (involving nearly 30,000 people) to answer one big question: Do these heart-helping drugs actually work, and are they safe?

Here is the breakdown in simple terms:

1. The Big Verdict: It's a Game Changer

Think of the heart failure patients as runners in a very tough marathon. The researchers found that giving these runners SGLT2 inhibitors was like giving them a super-charged energy drink.

• Fewer Deaths: The drugs helped people live longer. If you treated 100 people, about 10% fewer would die compared to those who didn't get the drug.
• Fewer Hospital Trips: This is the biggest win. The drugs acted like a "leak detector" and "pressure valve," stopping the heart from getting so overwhelmed that it needs emergency care. The study found a 26% drop in heart failure hospitalizations. That's like saving a huge number of people from being stuck in the hospital.
• Works for Everyone: It didn't matter if the heart was "weak" (low pumping power) or just "stiff" (high pumping power but stiff walls). The drug helped both types of engines.

2. The "Side Effect" Warning

Every tool has a trade-off. While the drugs were great for the heart, they had one annoying side effect.

Because these drugs work by flushing extra sugar out of the body through urine, it creates a sweet environment that bacteria love. This led to a 3.75 times higher chance of genital infections (like yeast infections).

• The Analogy: Imagine you are cleaning your house by opening all the windows to let fresh air in (great for the house!), but you also accidentally let in a few mosquitoes. The mosquitoes are annoying and need to be swatted (treated with simple medicine), but they aren't worth closing the windows and letting the house get stuffy.
• The Good News: The study found no increase in other scary side effects like kidney failure, dangerously low blood pressure, or diabetic ketoacidosis (a dangerous blood condition). In fact, the drugs seemed to protect the kidneys!

3. How They Did the Math

The researchers didn't just look at the 4 or 5 biggest, most famous studies. They dug deep into the "basement" of medical research to find 59 studies, including many smaller ones that others had ignored.

• Why this matters: It's like judging a movie not just by the reviews of the top 5 critics, but by reading 59 reviews from regular people, film students, and critics alike. This gave a much clearer, more complete picture of how the drug really works in the real world.

4. The "Engine" Details

• For Weak Engines (HFrEF): The drugs were amazing at preventing death.
• For Stiff Engines (HFpEF): The drugs were amazing at preventing hospital visits, even if they didn't drastically change the death rate in the short term.
• Quality of Life: People felt better. They could walk further and had less "breathlessness," which is like going from climbing a mountain with a backpack to walking up a gentle hill.

The Bottom Line

This massive review confirms that SGLT2 inhibitors are a "foundational" pillar for treating heart failure. They are like adding a new, essential safety feature to a car that was already good, making it safer, more reliable, and less likely to break down on the road.

The takeaway for patients: If you have heart failure, your doctor might suggest this drug. It will likely keep you out of the hospital and help you live longer. The main thing to watch out for is a higher chance of minor infections in private areas, which is usually easy to treat. The benefits to your heart engine far outweigh the annoyance of the mosquitoes.

Technical Summary

1. Problem Statement

While Sodium-Glucose Co-Transporter 2 (SGLT2) inhibitors are established as a cornerstone of heart failure (HF) therapy, previous meta-analyses have largely relied on a limited number of large, pivotal randomized controlled trials (RCTs). Consequently, the totality of randomized evidence—including smaller trials, specific safety outcomes, and data across the full spectrum of HF subtypes (HFrEF, HFpEF, HFmrEF)—has not been comprehensively synthesized. There is a need to evaluate the efficacy and safety of SGLT2 inhibitors using the complete body of available RCT data up to early 2026 to refine clinical understanding and guideline recommendations.

2. Methodology

• Study Design: A prospective systematic review and meta-analysis registered with PROSPERO (CRD420251167908), adhering to PRISMA 2020 guidelines.
• Data Sources: A comprehensive search of PubMed, Cochrane CENTRAL, ClinicalTrials.gov, and WHO ICTRP from inception through March 3, 2026.
• Eligibility Criteria:
  • Population: Adults (≥18 years) with clinical HF of any type (HFrEF, HFpEF, HFmrEF, or unspecified).
  • Intervention: Any SGLT2 inhibitor (e.g., empagliflozin, dapagliflozin, canagliflozin, sotagliflozin, etc.).
  • Comparator: Placebo or standard care.
  • Study Type: Randomized Controlled Trials (RCTs). Observational studies and single-arm trials were excluded.
• Data Extraction & Processing:
  • Included 59 RCTs (29,692 participants) after screening 6,239 records.
  • Utilized a rigorous 8-pass full-text screening pipeline and extracted data from full-text PDFs (approx. 175 fields per study).
  • Applied the Wan estimator for converting medians/IQRs to means/SDs where necessary.
• Statistical Analysis:
  • Models: Random-effects models using Restricted Maximum Likelihood (REML) estimation.
  • Effect Measures: Risk Ratios (RR) via Mantel-Haenszel method for binary outcomes; Mean Differences (MD) for continuous outcomes.
  • Confidence Intervals: Hartung-Knapp-Sidik-Jonkman (HKSJ) method for more robust coverage.
  • Heterogeneity: Assessed via Cochran Q, $I^2$, $\tau^2$, and 95% prediction intervals.
  • Bias & Certainty: Risk of Bias assessed via Cochrane RoB 2.0; Certainty of evidence graded using the GRADE framework.
  • Sensitivity: 12 filter-based sensitivity analyses, 4 alternative statistical models, leave-one-out analysis, and Trial Sequential Analysis (TSA).

3. Key Contributions

• Scope Expansion: This is the most comprehensive synthesis to date, incorporating 59 RCTs (29,692 participants) compared to previous meta-analyses that typically included only 4–10 pivotal trials. This includes data from smaller trials that were previously excluded.
• Granular Subgroup Analysis: Provided detailed stratification by HF subtype (HFrEF vs. HFpEF), drug class, blinding status, and follow-up duration, revealing nuanced interactions (e.g., mortality benefits driven primarily by HFrEF).
• Robust Statistical Rigor: Employed advanced statistical techniques (HKSJ CIs, TSA, trim-and-fill for publication bias) to address heterogeneity and small-study effects, offering a more reliable assessment of the "true" effect size.
• Safety Depth: Evaluated safety outcomes with greater statistical power due to the inclusion of smaller trials, providing definitive data on adverse events like genital infections, AKI, and DKA.

4. Key Results

Efficacy Outcomes

• All-Cause Mortality (ACM): Significant reduction with SGLT2 inhibitors (RR 0.90, 95% CI [0.83, 0.98], $p=0.016$).
  • Subgroup Note: Benefit was significant in HFrEF (RR 0.84) but not in HFpEF (RR 0.98).
  • Certainty: Low (downgraded for risk of bias and publication bias).
• Heart Failure Hospitalization (HFH): Significant reduction (RR 0.74, 95% CI [0.69, 0.79], $p<0.001$).
  • Consistent across all HF subtypes.
  • Certainty: Moderate.
• Composite Endpoint (CVD death + HFH): Significant reduction (RR 0.80, 95% CI [0.75, 0.85], $p<0.001$).
  • Certainty: High.
• Cardiovascular Death: Significant reduction (RR 0.86, 95% CI [0.76, 0.98], $p=0.031$).
• Quality of Life: KCCQ Total Symptom Score improved by 2.6 points (95% CI 1.2 to 4.0).
• Other Metrics: Modest weight reduction (-1.3 kg); no significant change in 6-minute walk distance or NT-proBNP.

Safety Outcomes

• Genital Infections: Significantly increased (RR 3.75, 95% CI [1.72, 8.19], $p=0.007$).
• Serious Adverse Events (SAE): Significantly reduced (RR 0.94, 95% CI [0.90, 0.99], $p=0.017$).
• No Significant Increases: Diabetic ketoacidosis, acute kidney injury (AKI), urinary tract infections, or hypotension.

Heterogeneity and Bias

• Heterogeneity: $I^2 = 0\%$ for primary outcomes (ACM and HFH), indicating high consistency across trials.
• Publication Bias: Detected for ACM (Egger's test $p=0.028$). Trim-and-fill analysis imputed 6 studies, adjusting the RR to 0.92, maintaining the direction of benefit but suggesting the original estimate may be slightly inflated by small-study effects.
• Risk of Bias: 36% of included trials were rated as "High Risk of Bias" (mostly due to open-label designs), which influenced the GRADE rating for ACM.

5. Significance and Implications

• Clinical Practice: The findings strongly support the inclusion of SGLT2 inhibitors as a foundational therapy across the entire spectrum of heart failure (HFrEF, HFpEF, and HFmrEF). The Number Needed to Treat (NNT) to prevent one HF hospitalization is 29, and for the composite endpoint is 24.
• Nuanced Mortality Benefit: The analysis clarifies that while HF hospitalization benefits are consistent across all ejection fractions, the mortality benefit is primarily driven by patients with reduced ejection fraction (HFrEF).
• Safety Profile: The therapy is confirmed to have a favorable safety profile regarding serious adverse events, with genital infections being the primary manageable risk. The lack of increased AKI or DKA risk is reassuring for broad application.
• Future Research: Highlights the need for longer follow-up to confirm mortality durability, dedicated trials for HFmrEF, and further investigation into newer agents and dual SGLT1/2 inhibitors.

Conclusion: This meta-analysis provides high-certainty evidence that SGLT2 inhibitors reduce HF hospitalizations and serious adverse events across all HF subtypes, with a mortality benefit specifically pronounced in HFrEF. Despite a low certainty rating for mortality due to bias concerns, the consistency of the data supports their use as a standard of care in heart failure management.