Skip to content
Sungho ParkRESEARCH & DISCOVERY
← All writing

Cardiovascular Research

Can Semaglutide Improve Cardiac Filling? CMR Insights From CAMEO-SEMA

Can a heart function better without a measurable improvement in ejection fraction? A prespecified cardiac magnetic resonance (CMR) substudy of the randomized CAMEO-SEMA trial offers an interesting example. In people with obesity-related heart failure with preserved ejection fraction (HFpEF), 52 weeks of semaglutide treatment was associated with improved left atrial deformation and more rapid early left ventricular filling compared with placebo.

Why look beyond ejection fraction?

In HFpEF, left ventricular ejection fraction (LVEF) remains relatively preserved even when filling pressures become abnormally high, particularly during exercise. Conventional volumetric measurements alone may therefore miss important aspects of cardiac dysfunction. Left atrial (LA) reservoir strain provides information about atrial deformation during ventricular systole, while the rate of early left ventricular (LV) filling describes how the ventricle refills after contraction.

The CAMEO-SEMA investigators used multiparametric CMR to ask whether previously observed hemodynamic benefits of semaglutide were accompanied by measurable changes in cardiac structure and function.

A randomized trial with paired CMR assessments

The substudy included 49 participants who completed both baseline and week-52 scans: 33 received semaglutide and 16 received placebo. All had symptomatic HFpEF and obesity (BMI at least 30 kg/m²). The median age was 67 years, and 28 participants were women. Semaglutide was administered once weekly, titrated to 2.4 mg, alongside usual care.

CMR was performed on a 1.5-T Siemens MAGNETOM Aera system. Cine imaging and feature tracking were used to evaluate atrial strain, ventricular volumes, myocardial mechanics, and cardiac geometry. Invasive exercise hemodynamics were assessed within one day of CMR, linking imaging findings with pulmonary artery wedge pressure (PAWP).

What changed after 52 weeks?

The prespecified primary CMR endpoint, LA reservoir strain, improved with semaglutide relative to placebo. Early LV filling dynamics also improved, alongside reductions in LV mass and favorable changes in right ventricular function.

MeasurementEstimated effect versus placeboP value
LA reservoir strain+5.7 percentage points0.0371
Early filling rate / LV EDV (eFR/EDV)+0.64 s−1<0.0001
LV mass−16.0 g<0.0001
RV ejection fraction+4.0 percentage points0.0049
Exercise PAWP at 20 W−11.0 mmHg<0.0001
Selected outcomes reported by Ghanbari et al. Values are Hodges–Lehmann estimates of the between-group difference in change from baseline, not individual patient changes. Original summary table compiled for this post.

Additional findings included improvement in LA conduit strain, LV global longitudinal strain magnitude, and RV free-wall strain. LA maximum volume was also lower. LVEF did not change significantly, emphasizing that a preserved ejection fraction does not rule out a treatment-associated change in other functional measurements.

How was LV filling measured with CMR?

This study did not estimate early filling from transmitral Doppler E-wave velocity or 4D flow MRI. Instead, the investigators constructed an LV volume-time curve from short-axis cine CMR images and calculated the slope of volume increase during a physiologically defined early-filling interval:

eFR = ΔLV volume / Δtime, followed by eFR/EDV = eFR / LV end-diastolic volume.

The distinction matters. Here, early filling rate (eFR) represents a window-averaged filling slope, rather than the maximum instantaneous frame-to-frame peak filling rate. Normalization to EDV also makes the metric less dependent on ventricular size. The study reported an improvement of 0.64 s−1 in eFR/EDV relative to placebo, with a corresponding improvement in the non-normalized eFR.

For CMR research, this highlights the value of extracting diastolic information from conventional cine images rather than relying on EF alone. It also means the reported metric should not be treated as interchangeable with Doppler E/A ratio or velocity-encoded flow parameters.

Do the imaging changes relate to exercise hemodynamics?

Participants with larger improvements in LA reservoir strain tended to show greater decreases in PAWP during 20-W exercise (Spearman r = −0.46). The corresponding association for eFR/EDV was stronger (r = −0.65). Changes in the two CMR functional measures were also correlated with one another (r = 0.59).

These associations support the physiological relevance of the imaging results, but correlation cannot establish which change caused another. Weight reduction, lower blood pressure, altered loading conditions, and cardiac remodeling may all contribute. The study cannot isolate a direct myocardial drug effect from these related pathways.

What should we be cautious about?

This was a relatively small, single-center CMR substudy involving 49 of the 81 randomized participants. Participants without paired scans were not included in the CMR analysis, and the cohort had limited racial diversity. The results specifically concern patients with obesity-related HFpEF and may not extend to all HFpEF populations.

The primary LA strain endpoint was prespecified, but numerous additional imaging and correlation analyses were examined without a multiple-comparison adjustment. Their findings are informative for understanding possible mechanisms and designing future research, rather than definitive proof that every measured feature will improve in broader clinical populations. The study also did not directly establish changes in intrinsic myocardial relaxation or long-term clinical outcomes through CMR.

What this means for cardiac MRI research

The most useful lesson is that cardiac response to therapy is multidimensional. LVEF was not significantly altered, yet atrial deformation, early diastolic filling kinetics, ventricular geometry, and exercise filling pressures showed favorable differences. Combining these measures may reveal aspects of therapeutic response that conventional EF measurements cannot capture on their own.

For related measurement context, explore the site’s cardiac structure and function CMR reference values and the discussion of E/A-based flow analysis in type 2 diabetes. These cover different populations and techniques and should not be interpreted as direct validation of the CAMEO-SEMA findings.

Original paper

Ghanbari F, Reddy YNV, Packer M, et al. Effects of semaglutide on cardiac structure and function in heart failure with preserved ejection fraction and obesity: a pre-specified cardiac magnetic resonance substudy of the CAMEO-SEMA trial. Nature Medicine. 2026. https://doi.org/10.1038/s41591-026-04787-8.

Editorial note: This is an independent research summary of an accepted article preview. It does not reproduce the publisher’s figures or prose. Findings should be checked against the final version of record when available. This article is not medical advice.