Moderate aortic stenosis has long been treated as a quiet way station on the road to severe valve disease, a stage worth watching but rarely worth acting on. A new review published in Cardiovascular Ultrasound argues that this view is overdue for revision. Drawing together registry data, longitudinal cohorts and imaging studies, the authors describe moderate aortic stenosis as a heterogeneous and clinically meaningful entity that carries substantial morbidity and mortality even before the valve reaches the severe range, and they lay out a multimodality imaging framework designed to find the patients who are quietly at highest risk.
The scale of the problem is considerable. Registry data suggest that moderate aortic stenosis accounts for 30 to 45 percent of clinically relevant aortic stenosis in elderly cohorts. Yet the true prevalence may be lower than these figures imply, because standard transthoracic echocardiography, the workhorse of diagnosis, is prone to measurement errors that can shift patients between severity categories. In a multicenter study cited in the review, up to 23 percent of patients labeled severe by echocardiography were reclassified as moderate when computed tomography-derived aortic valve area exceeded 1.0 square centimeter, and prospective studies confirm that up to one-quarter of patients initially labeled moderate are re-stratified after multimodality testing.
The technical pitfalls are specific and well documented. Peak jet velocity can be underestimated if continuous-wave Doppler recordings are not systematically obtained from all acoustic windows, including apical, right parasternal and suprasternal views. Mismeasurement of the left ventricular outflow tract diameter, often caused by assuming a circular geometry for an anatomically elliptical structure, propagates cubically in the continuity equation and can substantially bias the calculated valve area. When flow acceleration in the outflow tract exceeds 1 meter per second, the simplified Bernoulli equation no longer accurately represents the transvalvular pressure gradient, and the complete formulation must be applied to avoid systematic overestimation. Discordance between valve area and mean gradient is frequent in the mild-to-moderate spectrum, and the review proposes a structured stepwise reassessment, including stroke volume index, transaortic flow rate and, where uncertainty persists, CT-derived valve area, before any reclassification of severity.
Physiology adds further layers of confusion. Classical low-flow states, defined as a stroke volume index below 35 milliliters per square meter, can arise from concentric remodeling, small left ventricular cavities, tachyarrhythmias, conduction delays or increased afterload from vascular stiffness, and in these settings Doppler gradients may underestimate severity. The transaortic flow rate, calculated as stroke volume divided by left ventricular ejection time, has emerged as a more robust marker: in a large multicenter cohort, a flow rate below 218 milliliters per second was independently associated with a twofold higher risk of mortality and heart failure hospitalization, outperforming stroke volume index. Pressure recovery, more pronounced in patients with small aortic roots, can make Doppler gradients overestimate true severity, while the acceleration time to ejection time ratio offers a simple, reproducible marker that may identify concealed low-flow physiology at higher risk of rapid progression.
Computed tomography and cardiac magnetic resonance fill the gaps that Doppler leaves open. CT provides anatomical planimetry of the valve area and aortic valve calcium scoring, both independent of flow conditions, with sex-specific calcium thresholds enhancing diagnostic accuracy. Refined graduated calcium tiers tailored to the mild-to-moderate spectrum have been proposed, acknowledging that risk within the moderate category is graded rather than binary. CMR, by contrast, excels at flow quantification and myocardial characterization, detecting focal replacement fibrosis on late gadolinium enhancement and diffuse interstitial fibrosis on T1 mapping and extracellular volume quantification, findings independently associated with adverse outcomes. The review is careful to note the caveats: in cardiac amyloidosis or bicuspid valves, calcification may not parallel hemodynamic severity, so calcium scores must always be interpreted alongside morphology and hemodynamic data.
Molecular imaging extends this paradigm further. Fluorine-18 sodium fluoride positron emission tomography detects active microcalcification at the molecular level, before macroscopic calcium becomes visible on CT, and increased uptake predicts calcium progression and correlates with accelerated hemodynamic deterioration. Serial CT calcium scoring identifies fast progressors, patients whose valves deteriorate rapidly and who may benefit from personalized surveillance. The review, however, is explicit that PET imaging with sodium fluoride and emerging tracers such as fibroblast activation protein inhibitors remain primarily research tools, not part of routine clinical decision-making, as does four-dimensional flow CMR.
Perhaps the most conceptually important shift is the move from valve-centric to whole-heart staging. Aortic stenosis is increasingly framed as a systemic disease, a continuum of progressive damage involving the left ventricle, left atrium, right ventricle and pulmonary circulation. In an extensive multicenter study, Tastet and colleagues extended this staging to asymptomatic patients, including those with moderate disease, showing that progressive stages of cardiac damage were strongly and independently associated with adverse outcomes irrespective of traditional severity definitions. Crucially, in moderate aortic stenosis, higher stages identified subsets whose outcome trajectories approached those of severe disease. Multi-chamber strain imaging reinforces this picture: abnormalities in left atrial reservoir strain and right ventricular free-wall strain are often detectable despite preserved left ventricular global longitudinal strain, and patients with abnormal values show progressively worse outcomes. The authors stress that this framework rests on cross-sectional, associative data and does not establish causality, and that there is currently no evidence that valve replacement reverses fibrosis markers or improves prognosis on that basis.
The prognostic numbers are sobering. In an extensive Australian registry of 3,315 patients, the five-year mortality rate in moderate aortic stenosis reached 56 percent, comparable to that of severe disease. In European studies of moderate aortic stenosis with left ventricular dysfunction, 61 percent of patients reached a composite endpoint of death, valve replacement or heart failure hospitalization within four years. Meta-analytic data involving more than 35,000 patients indicate that approximately 15 to 20 percent progress to severe disease or experience major cardiovascular events within five years. Elevated natriuretic peptides, high-sensitivity troponin and fibrosis-related biomarkers such as galectin-3 and soluble ST2 add independent prognostic information, and the coexistence of cardiac amyloidosis, which is not rare, warrants systematic evaluation when symptoms or remodeling seem disproportionate to valve hemodynamics.
Artificial intelligence is entering the workflow at a pragmatic level. Deep-learning models can automatically quantify peak velocity, mean gradient, outflow tract dimensions and valve area on routine echocardiography with strong agreement with expert readers, and automated CT algorithms perform annular sizing and root segmentation with comparable accuracy while reducing reporting time. Because outflow tract errors propagate cubically in the continuity equation, models that standardize its measurement across vendors may stabilize valve area calculations. Machine-learning models integrating imaging with demographics and laboratory data may eventually predict progression and guide individualized surveillance intervals. But the review cautions that dataset bias, limited generalizability, black-box outputs and a predominantly retrospective evidence base mean AI approaches in moderate aortic stenosis remain investigational and should complement, not replace, expert assessment.
On management, guideline-directed surveillance remains the standard: echocardiographic follow-up every one to two years in moderate disease, shortened when high-risk features such as rapid velocity increase, high calcium burden, abnormal strain or elevated natriuretic peptides are present. Early valve replacement is not currently indicated for moderate stenosis alone, and no drug has been shown to alter disease progression, with statin trials having failed to demonstrate benefit. The TAVR-UNLOAD randomized trial of 178 patients with moderate stenosis and heart failure with reduced ejection fraction found no significant reduction in mortality, disabling stroke or heart failure hospitalizations over a median 23 months, though quality of life improved; high comorbidity burden and a 43 percent crossover rate limited the result. A meta-analysis of seven studies and 4,827 patients associated early valve replacement with a 48 percent reduction in all-cause mortality, and two ongoing randomized trials, PROGRESS and EXPAND TAVR II, will test whether early transcatheter intervention improves outcomes in selected high-risk phenotypes. Until those results arrive, the authors conclude, cardiologists should deploy every multimodality imaging tool available to identify the patients in whom earlier treatment of moderate aortic stenosis is most likely to improve long-term outcomes.
Subject of Research: Diagnostic pitfalls and multimodality imaging in moderate aortic stenosis
Article Title: Moderate aortic stenosis: diagnostic pitfalls and the role of multimodality imaging in risk stratification and clinical management
Article References: Santangelo, G., Maloberti, A., Tognola, C., Monticelli, M., Montalto, C., Barbieri, A., Moreo, A., Giannattasio, C., Carugo, S., & Faggiano, P. (2026). Moderate aortic stenosis: diagnostic pitfalls and the role of multimodality imaging in risk stratification and clinical management. Cardiovascular Ultrasound, 24(1), Article 19. https://doi.org/10.1186/s12947-026-00381-7
Image Credits: AI Generated
DOI: 10.1186/s12947-026-00381-7
Keywords: aortic stenosis, echocardiography, computed tomography, cardiac magnetic resonance, aortic valve calcium, artificial intelligence, risk stratification, TAVR, heart failure, 18F-sodium fluoride PET, guidelines, valve replacement
News Source: Ophelia Keating. (October 28, 2026). Moderate aortic stenosis is deadlier than thought, and imaging may unmask it. Scienmag.
