Chinese Ophthalmic Nurses Negotiate Trust in Generative AI

A review of three randomized controlled trials involving 455 adults with heart failure found that sodium-glucose cotransporter-2 (SGLT2) inhibitors, when added to standard heart failure treatment, were linked to small reductions in functional mitral regurgitation. This condition occurs when blood leaks backward through the mitral valve because changes in the heart affect how the valve closes. Across the three trials, these medicines reduced the effective regurgitant orifice area, which measures the opening through which blood leaks, by an average of 0.086 square centimeters. They also reduced the amount of blood leaking backward through the valve by an average of 10.87 milliliters compared with control treatment. Both combined results favored SGLT2 inhibitors. However, the findings were sensitive to which studies were included, differed considerably across studies and were based on evidence judged to be of low certainty. The results suggest that these medicines may reduce functional valve leakage, but they do not prove that the drugs directly treat the valve problem.

Functional mitral regurgitation, also called secondary mitral regurgitation, happens when disease or changes in the shape of the heart's left ventricle affect how the mitral valve works. The problem is not caused by direct damage to the valve's flaps. Instead, the forces that pull the flaps apart become unbalanced with the forces that bring them together to close the valve. The small muscles that help control the valve may move out of place, the ring around the valve may widen, and the ventricle may become less able to close the valve firmly. This condition affects about one-third of people with heart failure and is linked to a poor outlook. SGLT2 inhibitors are already used to treat heart failure and have been linked to changes that help the heart's shape and structure move back toward normal. Researchers therefore investigated whether these medicines might also reduce functional mitral regurgitation.

The researchers searched PubMed, Embase, Scopus, Web of Science and the Cochrane Central Register of Controlled Trials, covering records from when each database began until July 21, 2026. They screened 745 records, examined four full research reports and included three randomized trials. The studies took place in South Korea, China and Egypt between 2024 and 2025. At the start, the trials had randomly assigned 472 participants to treatment groups. A total of 455 completed follow-up and were included in the combined analysis. All participants had long-term heart failure and functional mitral regurgitation despite receiving stable background treatment. This treatment included some standard heart failure medicines, such as beta-blockers, drugs that affect the renin-angiotensin system and mineralocorticoid receptor antagonists.

The three trials had important differences in their design and the length of follow-up. EFFORT was a six-center, double-blind trial in South Korea that compared ertugliflozin with a placebo, a treatment with no active study drug. It included 128 participants and followed them for 12 months. The DEFORM trial in China included 104 patients and compared dapagliflozin at a dose of 10 mg once a day with standard heart failure treatment alone. It measured results after three months. The third trial, conducted in Egypt, included 240 patients and made the same comparison using dapagliflozin 10 mg once a day, with follow-up lasting six months. Two studies were open-label, meaning participants and treating teams knew which treatment was given. However, the people assessing heart ultrasound results did not know the treatment assignments. EFFORT used a placebo and kept treatment assignments hidden. Across the combined group, the average age was 61.5 years, and 70.3% of participants were male.

For the main heart imaging results, adding an SGLT2 inhibitor reduced the effective regurgitant orifice area by an average of 0.086 square centimeters compared with control treatment. The mean difference was −0.086 square centimeters, with a 95% confidence interval from −0.150 to −0.022 and a p-value of 0.0089. The 95% confidence interval shows the range of values that fit the data under the statistical method used. For the amount of blood leaking backward through the mitral valve, the mean difference was −10.87 milliliters, with a 95% confidence interval from −17.32 to −4.43 and a p-value of 0.0009. These changes were small in clinical terms. The researchers also found substantial differences between the three trials. The I² value, a measure of how much study results differ beyond what would be expected from chance alone, was 72.5% for effective regurgitant orifice area and 63.5% for regurgitant volume. The reviewers examined these differences alongside variations in participant characteristics, treatments, follow-up periods and the severity of valve leakage at the start of each study.

The researchers carried out sensitivity analyses by removing one study at a time to see how much each trial affected the results. Removing the Abou Warda study reduced the differences between studies for effective regurgitant orifice area, with I² falling from 72.5% to 46.9%. This suggested that the larger, open-label trial was a major source of variation. However, removing the placebo-controlled EFFORT trial meant that neither of the two main outcomes remained statistically significant. Without EFFORT, the estimated difference in effective regurgitant orifice area was still negative at −0.103 square centimeters, but the 95% confidence interval ranged from −0.236 to +0.029, and the p-value was 0.1251. For regurgitant volume, the estimated difference remained negative at −12.03 milliliters, but the p-value increased to 0.0801. These findings led the researchers to rate the evidence lower because the study results were inconsistent and the estimates were not precise enough.

The results for major health events appeared stronger in numerical terms, but they were even more affected by the small amount of available data. The combined analysis linked SGLT2 inhibitor treatment to a 55% lower relative risk of hospitalization for heart failure. The risk ratio was 0.45, with a 95% confidence interval from 0.28 to 0.73. It also found a 74% lower relative risk of death from cardiovascular causes, with a risk ratio of 0.26 and a 95% confidence interval from 0.12 to 0.58. Neither analysis found clear differences between the studies for these outcomes. However, the researchers rated the evidence as low certainty for heart failure hospitalization and very low certainty for cardiovascular death. Based on the event rates in the control groups, the review estimated that hospitalizations for heart failure would fall from 177 to 80 per 1,000 patients, while deaths from cardiovascular causes would fall from 97 to 25 per 1,000 patients. These figures are estimates from the combined results of just 455 participants, not definitive proof of how much the treatment reduces these risks.

The review warned that the estimates for these health events were driven mainly by one study and suggested much larger benefits than those found in much bigger trials of SGLT2 inhibitors. For comparison, the authors cited broader evidence from trials involving more than 20,000 randomly assigned patients. Those studies reported approximate hazard ratios of 0.72 for a first hospitalization for heart failure and 0.87 for cardiovascular death. A hazard ratio compares how often an event happens over time between groups. The risk ratios of 0.45 and 0.26 in the current review suggest much larger relative benefits. In particular, the estimated reduction in cardiovascular death was nearly fourfold, based on only 26 deaths. The authors said that small studies, very few events, unblinded assessment of secondary outcomes and possible selective reporting were more likely explanations for these unusually large estimates than a true increase in the treatment's benefits. The included trials were mainly designed to study how the drugs might work, rather than to provide reliable evidence about major health outcomes.

Measurements of how the heart worked suggested that the drugs might affect the heart's main pumping chamber and nearby structures rather than repair the mitral valve's flaps directly. In the Egyptian trial, reduced valve leakage occurred alongside decreases in left ventricular mass and volume after adjusting for body size. However, EFFORT and DEFORM found no significant difference between treatment groups in left ventricular volume, even though valve leakage decreased. Both studies reported improvements in measures linked to the heart's ability to contract, the pressure inside the heart as it fills with blood, or the shape of the upper left chamber. These measures included global longitudinal strain, ejection fraction, E/e′ and left atrial volume index. The authors suggested that stronger pumping, lower filling pressures or less stretching of the left atrium might help the valve flaps meet and close more effectively before measurable reductions in ventricular volume occur. They also considered reducing pressure and strain on the left atrium especially relevant because 26.3% of the combined group had atrial fibrillation, an irregular heart rhythm. These ideas are possible explanations rather than proven mechanisms because none of the trials tested whether any specific change in heart function caused the reduction in valve leakage.

Several limitations make the findings harder to interpret and apply to other patients. Reducing preload, the amount of blood returning to the heart before it contracts, can temporarily change the shape of the ventricle and the pressure across the mitral valve. These changes can reduce measured valve leakage without showing that lasting structural improvement has occurred. The DEFORM trial measured heart imaging results after just three months, so it could not distinguish a temporary change in blood flow and pressure from lasting changes in heart structure. In that study, participants taking dapagliflozin were more likely to have their background furosemide dose reduced or the medicine stopped altogether. This meant that the treatment groups did not have equal fluid levels at follow-up. The severity of valve leakage at the start also differed across the trials. The average effective regurgitant orifice area was 0.20 square centimeters in EFFORT, compared with 0.43–0.45 square centimeters in the Egyptian trial. The studies excluded patients with preserved ejection fraction, which means the heart pumps out a normal proportion of the blood it contains. EFFORT also excluded patients whose ejection fraction was below 35%. These exclusions limit what can be concluded about people with severe heart muscle weakness and those whose mitral regurgitation is mainly related to changes in the left atrium. Only dapagliflozin and ertugliflozin were studied. Reporting of side effects varied too much across the trials to allow a reliable combined safety analysis. The researchers also had only data from whole studies, not individual participants, so they could not reliably examine how effects differed across patient subgroups. Larger placebo-controlled trials with consistent conditions for measuring heart function, carefully confirmed health outcomes and planned valve-related outcomes are needed before these combined findings can be considered confirmation of a treatment benefit.

Subject of Research: How SGLT2 inhibitors affect functional mitral regurgitation in adults with heart failure

Article Title: Efficacy of sodium-glucose cotransporter-2 inhibitors on functional mitral regurgitation associated with heart failure: a systematic review and meta-analysis of randomized controlled trials

Article References: Kajo, C., Alaraibi, A. J., Almutairi, O. T., Alaraibi, Y. H., Alibrahim, A., Almutawah, Y. K. A., Alasfar, S. W. S., Ali, M. W., Hajjiri, A., Alqaryuti, M., Alfehaid, A., & Arekat, D. (2026). Efficacy of sodium-glucose cotransporter-2 inhibitors on functional mitral regurgitation associated with heart failure: a systematic review and meta-analysis of randomized controlled trials. Frontiers in Cardiovascular Medicine, 13, Article 1976876. https://doi.org/10.3389/fcvm.2026.1976876

Image Credits: AI Generated

DOI: 10.3389/fcvm.2026.1976876

Keywords: SGLT2 inhibitors, functional mitral regurgitation, heart failure, mitral valve, randomized controlled trials, meta-analysis, echocardiography, cardiovascular death, heart failure hospitalization, certainty of evidence, efficacy, sodium-glucose